# Seminars

Date | Lecturer | Title | Location |
---|---|---|---|

15:00 - 17:15 | E. Flores Tames (Complexity Science Hub Vienna) | Supply chain ID with payment data | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | N. Reisz (Complexity Science Hub Vienna) | Polya urn / CCC / APS | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | S. Lindner (University of Vienna) | Non extensive entropies / Magnetic coin model | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | C. Puchhammer (University of Vienna) | Embrionic mouse development | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | M. Ferreira Goncalves (Complexity Science Hub Vienna) | Knowledge as recombination | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | C. Diem (International Institute for Applied Systems Analysis) | Multilayer "DebtRank" | tba |

15:00 - 17:15 | B. Corominas-Murtra (Institute of Science and Technology Austria) | tba | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | A. Di Natale (Medical University of Vienna) | Colexification networks | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | J. Reddish (Complexity Science Hub Vienna) | Introduction to Seshat Global History Databank | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | V. Hemmelmayr (Vienna University of Economics and Business) | Location and routing decisions for a collaborative recycling network Abstract This talk presents an application in collaborative recycling efforts for non-profit agencies. The work is motivated by a project with a network of hunger relief agencies (e.g., food pantries, soup kitchens and shelters) focusing on collaborative approaches to address their cardboard recycling challenges collectively. The problem is modeled as a periodic location routing problem with operational choice. A metaheuristic is compared to the exact solution of the problem. The relative difficulty introduced with each decision is examined. Computational results on synthetic instances as well as a case study based on data from the network will be presented. In this novel setting, we evaluate collaboration in terms of participation levels and cost impact. These insights can be generalized to other networks of organizations that may consider pooling resources. | CSHV, 8., Josefstädter Str. 39, room 201 |

15:00 - 17:15 | W. Schueller (Complexity Science Hub Vienna) | Active control of complexity growth in a multi-agent model | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-10-11 15:00 - 17:15 | F. Neffke (Harvard Kennedy School) | Diffusion and coordination of complex knowledge Abstract The talk will discuss knowledge diffusion and knowledge coordination in modern economies. Knowledge, or better know-how, is an important determinant of economic prosperity and growth. With a growing global body of know-how, but more or less constant capacity of individuals to master it, economies face the problem that know-how becomes more and more distributed across people and teams of people. Tapping and coordinating this know-how is therefore crucial to the success of complex economies. The talk will present a number of examples of how large data sets, such as social security, cell phone and payment card data, can be used to explore the diffusion and coordination of know-how, highlighting the role of migration, foreign investments, business travel and cities in a globalizing economy.
<p></p>
Bio: Frank Neffke is the Research Director of the Growth Lab at the Harvard Kennedy School. His research focuses on economic transformation and growth. He has written on a variety of topics, such as structural transformation and new growth paths in regional economies, economic complexity, division of labor and teams, the consequences of job displacement and the future of work. Before joining the Growth Lab, Frank worked as an assistant professor at the Erasmus School of Economics in Rotterdam, The Netherlands. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-06-28 15:00 - 17:15 | T. Pham (Medical University of Vienna) | Opinion formation and Structural Balance on Coevolving Signed Networks Abstract We study opinion dynamics on a network of friendships and enmities, where agents' opinions and the relationships among them coevolve, although on separate timescales. We model the
agents opinions and their relationships as binary spins and coupling constants, respectively.
The dynamics of both spins and couplings are implemented via the Metropolis algorithm, assuming they are coupled to heat baths at different temperatures. For this coevolutionary dynamics, at equilibrium, we find a continuous transition from an ordered phase, where agents are partitioned into two antagonistic groups to a mixed phase, where the bipartition disappears and opinions are randomly distributed amongst agents. If only the agents'
relationships change over time while the opinions remain unchanged, then, in addition to the disappearance of the network bipartition, we observe a spin-glass-like transition from a unique network configuration to multiple possible configurations. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-06-14 15:00 - 17:15 | S. Schweighofer (Medical University of Vienna) | A Balance Model of Opinion Hyperpolarization Abstract Polarization is threatening the stability of democratic societies. Until
now, research has focused on the extremeness aspect of polarization, and
failed to explain the correlation between different polarized issues, i.e., \emph{hyperpolarization}. We develop an opinion dynamics model based on psychological balance theory that, for the first time, is able to generate
hyperpolarization, and to explain the link between affective and opinion
polarization.
We validate the predictions of this model on empirical data from the
2016 National Election Survey.
| CSHV, 8., Josefstädter Str. 39, room 201 |

2019-06-07 15:00 - 17:15 | O. Saukh (Graz University of Technology and Complexity Science Hub Vienna) | Quantle: Fair and Honest Presentation Coach in Your Pocket Abstract Great public speakers are made, not born. Practicing a presentation in front of colleagues is common practice and results in a set of subjective judgements what could be improved. In this paper we describe the design and implementation of a mobile app which estimates the quality of speaker’s delivery in real time in a fair, repeatable and privacy-preserving way. Quantle estimates the speaker’s pace in terms of the number of syllables, words and clauses, computes pitch and duration of pauses. The basic parameters are then used to estimate the talk complexity based on readability scores from the literature to help the speaker adjust his delivery to the target audience. In contrast to speech-to-text-based methods used to implement a digital presentation coach, Quantle does processing locally in real time and works in the flight mode. This design has three implications: (1) Quantle does not interfere with the surrounding hardware, (2) it is power-aware, since 95.2 % of the energy used by the app on iPhone 6 is spent to operate the built-in microphone and the screen, and (3) audio data and processing results are not shared with a third party therewith preserving speaker’s privacy.
We evaluate Quantle on artificial, online and live data. We artificially modify an audio sample by changing the volume, speed, tempo, pitch and noise level to test robustness of Quantle and its performance limits. We then test Quantle on 1017 TED talks held in English and compare computed features to those extracted from the available transcript processed by online text evaluation services. Quantle estimates of syllable and word counts are 85.4 % and 82.8 % accurate, and pitch is over 90 % accurate. We use the outcome of this study to extract typical ranges for each vocal characteristic. We then use Quantle on live data at a social event, and as a tool for speakers to track their delivery when rehearsing a talk. Our results confirm that Quantle is robust to different noise levels, varying distances from the sound source, phone orientation, and achieves comparable performance to speech-to-text methods. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-05-24 15:00 - 17:15 | J. Korbel (Medical University of Vienna) | Predicting collapse of networked systems without knowing the network Abstract Networked systems are ubiquitously present in the real world – food webs, inter-bank markets or communication networks,
for instance. Prediction of collapses in networked systems is an extremely important, still almost impossible task. The
main obstacle is typically the enormous size of the network resulting in extremely large amount of information necessary to describe its structure. Moreover, in many systems the network structure remains hidden. Based on the corollary of the famous Perron-Frobenius theorem called eigenvector quantization, we show that for a broad class of networked systems it is possible
to detect the last stage before the crash without knowing the network structure and consequently to predict the collapse of the whole system. | CSHV, 8., Josefstädter Str. 39, Salon |

2019-05-17 15:00 - 17:15 | C. Poellabauer (University of Notre Dame) | Speech as Barometer of the Brain Abstract Recent projections indicate that the number of connected Internet-of-Things (IoT) devices, sensors, and actuators will pass 46 billion in 2021 and people will interact with these devices on a daily basis, e.g., to monitor and control our health, home, connected vehicles, security systems, social interactions, and almost every other aspect of our daily activities. As one of the most natural ways of communication, speech has recently found a rapidly increasing interest as primary mode of interaction between humans and IoT devices. At the same time, recent research has shown that there are clear links between the neurological and mental wellness of an individual and patterns in the individual's speech. Timely detection of such impairments can help improve the user-machine interactions (e.g., adapt the computing systems to the mental or cognitive conditions of the user) and prevent problems (such as operation of safety-critical equipment by users lacking appropriate cognitive fitness) before they cause damage to humans and machines. In this talk, I will discuss several challenges in continuous and non-intrusive speech assessment and present our ongoing research efforts in several specific case studies.
Bio: Christian Poellabauer received his Dipl. Ing. degree from the Vienna University of Technology, Austria in 1998 and the Ph.D. degree from the Georgia Institute of Technology, Atlanta, GA in 2004, both in Computer Science. He is an Associate Professor in the Department of Computer Science and Engineering at the University of Notre Dame and was awarded a 2019 Fulbright Scholar grant to pursue his academic interests at TU Graz. His research interests are in the areas of wireless sensor networks, mobile computing, ad-hoc and vehicular networks, pervasive computing, and mobile healthcare systems. He has published over 150 scientific contributions in these areas and he has co-authored a textbook on Wireless Sensor Networks. His research has received funding through the National Science Foundation, the National Institutes of Health, Department of Education, Department of Defense, IBM, Intel, Toyota, Ford Research, Motorola Labs, National Geographic, the National Football League, GE Health, and several other foundations and businesses. He received the Outstanding Dissertation Award from Georgia Tech and an NSF CAREER Award in 2006. He is a senior member of ACM and IEEE. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-05-10 15:00 - 17:15 | N. Haug (Medical University of Vienna) | History-dependent Modeling of Patient Health Trajectories Abstract Multimorbidity, the co-occurrence of two or more chronic diseases such as diabetes, obesity or cardiovascular diseases in one patient, is a frequent phenomenon.
To make care more efficient, it is of relevance to understand how different diseases condition each other over the life course of a patient.
However, most of our current knowledge on such patient careers is either confined to narrow time spans or specific (sets of) diseases.
Here, we present a population-wide analysis of long-term patient trajectories by clustering them according to their disease history observed over 17 years.
When patients acquire new diseases, their cluster assignment might change.
A health trajectory can then be described by a temporal sequence of disease clusters.
From these cluster transitions we construct an age-dependent multiplex network of disease clusters.
Random walks on this multiplex network provide a more precise model for the time evolution of multimorbid health states when compared to models that cluster patients based on single diseases.
We find that for elderly patients the cluster network consists of two different regions, one for clusters with low and one with high in-hospital mortality.
Our results can be used to identify the crucial events that potentially determine the future disease trajectory of a patient. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-05-03 15:00 - 17:15 | M. Pellert (Medical University of Vienna) | Emotional expression dynamics in social media Abstract We analyzed a dataset of general social media text-based expression: 22 Million status updates donated by 153727 Facebook users. By applying a lexicon-based approach, we quantified emotions in each post in terms of valence, the degree of pleasure associated with an emotional experience, and arousal, the level of activity induced by the emotional experience. We further replicated our analysis with the unsupervised VADER method and supervised valence and arousal methods specific for Facebook. We analyzed how the emotions of a post predict the emotional content of the next post by the same user as a function of time between posts ?t. It can be observed that emotional expression relaxes quickly but not instantly, as changes in short timescales are directed towards the baseline but do not fully converge. We captured this dynamics through a dynamic equation model following an agent-based modelling framework for emotions in online interaction. This way we modelled emotion eigendynamics as an example of an Ornstein–Uhlenbeck process. We fitted the solution of this model against both datasets through nonlinear regression and came up with three main results: i) the presence of emotional expression indicates emotion regulation, evidencing a correction of emotional intensity as soon as a message is written, ii) valence and arousal decay exponentially towards their baselines, and iii) the baseline of valence is above its midpoint but the baseline of arousal is slightly below its midpoint (on a 1-9 scale). The above results are in line with well-established dynamics of emotions from previous research in psychology, but our analysis has several advantages. We observe a much longer period (for Facebook on average 534.36 days) over a larger population than the largest previous works with self-reports. Additionally, we capture general emotion dynamics rather than changes due to emotion labelling: This enables us to quantify individual dynamics that inform future computational models and analyses of emotional expression in social media. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-04-12 15:00 - 17:15 | L. Zavojanni (Medical University of Vienna) | Prediction of human behaviour in the multiplayer online game Pardus Abstract Human society relates through different sorts of interactions that are not independent of each other. The question arises whether these interdependencies can be used to predict the dynamical behaviour of such relationships. Here we look for empirical evidence that the dynamic behaviour of a given layer of a social system is related to the behaviour of other layers. Succeeding in this endeavour would make possible to improve predictions. But it would also improve the understanding of humanity as a set of interacting elements and the social dynamics between these elements.
The Massive Multiplayer Online Game (MMOG) Pardus is the source of data. In this game, players live in a virtual, futuristic universe where they interact with each other in a multitude of ways to achieve their self-posed goals. Such system represents a complete human society.
If the interaction between two people is represented as a link between two nodes, the whole set of human interaction can be described as a multiplex network. In such a system, layers represent different sorts of interactions.
One basic phenomena in a network is the formation and disappearance of links over time. A key question is whether it is possible to predict link appearances in one layer of the multiplex system using knowledge about the other layers. To answer this, the well-known problem of link prediction is generalized by extending it to the multiplex scenario represented by our social system. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-04-05 15:00 - 17:15 | B. Trpin (University of Salzburg & University of Ljubljana) | Belief updating in untrustworthy circumstances Abstract Lying is traditionally defined as stating something that is believed to be false with the intention that the other person believes it to be true. However, there are cases where we believe something to be false only to an extent and we can nevertheless lie about it. We call this partial lying. We provide an epistemological analysis of these situations and investigate how the severity of partial lying may be determined, and how partial lies can be classified. To make the investigation more realistic, we also study how much epistemic damage an agent suffers depending on her level of trust that she invests in the liar and the degree of belief in the falsehood of the liar's statements. The results are demonstrated through computer simulations of an arguably rational Bayesian agent who is trying to determine how biased a coin is while observing the coin tosses and listening to an (un)trustworthy person's misleading predictions about the outcomes. Our results provide an interesting insight at the intersection of epistemology and ethics, namely that in the longer term partial lies lead to more epistemic damage than outright lies. However, if the liar only lies sometimes, then categorical lies are more epistemically damaging as the liar is considered to be more trustworthy. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-03-29 15:00 - 17:15 | K. Lerman (University of Southern California) | Simpson’s Paradox in Social Data Abstract Socialdata often comes from a heterogeneous population composed of non-randomly sampled subgroups, each with different characteristics and behaviors. A population-level trend in the aggregate data may disappear or reverse itself when the same data is disaggregated into its underlying subgroups. This effect, known as Simpson’s paradox, confounds analyses of social data, including inference of trends and causal effects.
I illustrate the problem with several examples showing how the paradox can distort conclusions of analysis, and describe recent algorithmic efforts to address this problem. Our algorithm systematically disaggregates data to identify subgroups whose behavior deviates significantly from the rest of the population. The method allows us to leverage Simpson’s paradox to uncover interesting patterns in real-world social data, such as Q&A site Stack Exchange and online learning platforms Khan Academy and Duolingo.
<p></p>
<b>Kristina Lerman</b> is a Principal Scientist at the University of Southern California Information Sciences Institute and holds a joint appointment as a Research Associate Professor in the USC Computer Science Department. Trained as a physicist, she now applies network analysis and machine learning to problems in computational social science, including crowdsourcing, social network and social media analysis. Her recent work on modeling and understanding cognitive biases in social networks has been covered by the Washington Post, Wall Street Journal, and MIT Tech Review.
| CSHV, 8., Josefstädter Str. 39, room 201 |

2019-03-22 15:00 - 17:15 | C. Matzhold (University of Vienna) | How interdisciplinary research promotes data-driven analysis Abstract The amount of medical data generated in healthcare has exponentially increased in recent years. To gain a quantitative understanding of the complex processes involved in healthcare, we need new scientific methods to adequately investigate the heterogeneous and dynamic data they produce. Data-driven analysis, when applied in a clinical context, has the aim to turn complex medical data records into knowledge on how to prevent or treat diseases more effectively. This approach requires the combination of data-related, methodological expertise with in-depth know-how of the involved clinical or medical domain. More precisely, it requires interdisciplinary collaboration, since data analysis competencies needs to be consistent with medical domain expertise. However, successful interdisciplinary research remains a great challenge. In order to transform medical data into knowledge, we need to gain new insights into factors that support interdisciplinary interaction and the emergence of cognitive processes through which disciplinary concepts and methods are integrated. Therefore, this master thesis has two objectives. First, to carry out a data-driven analysis by means of an interdisciplinary research team and, second, to combine our practical experience with theoretical knowledge about cognitive processes to gain knowledge about interdisciplinary research. In terms of methodological data analysis approaches, we applied multiple logistic regression to analyse the dosage-dependent occurrence of osteoporosis in statin patients. This study reveals that there is a highly non-trivial dependence of statin dosage with the odds of osteoporosis. To the best of our knowledge, this is the first study which shows that it is important to consider both potency and dosages when investigating the relationship of osteoporosis and statin therapy. Our results show that the diagnosis of osteoporosis is underrepresented in low-dose and overrepresented in high-dose statin treatment. The second high-level objective was to show how interdisciplinarity promotes data-driven analysis and to gain insights into factors that support the interaction process between researchers from different disciplines and the emergence of a cognitive process among them. We made use of contemporary cognitive science theories and Bourdieu’s theory of practice to get an alternative perspective on interdisciplinarity. Based on these theoretical insights combined with our own research experience, we identified body—mind--environment actors that we consider important for a successful research process: personal/habitus-related qualities, environmental/field-related properties, a shared study logic and time. In summary, this work demonstrates how data-driven analysis is promoted by interdisciplinary research. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-03-15 15:00 - 17:15 | P. Klimek (Medical University of Vienna) | Quantifying economic resilience from input--output susceptibility to improve predictions of economic growth and recovery Abstract Modern macroeconomic theories were unable to foresee the last Great Recession and could neither predict its prolonged duration nor the recovery rate. They are based on supply--demand equilibria that do not exist during recessionary shocks. Here we focus on resilience as a nonequilibrium property of networked production systems and develop a linear response theory for input--output economics. By calibrating the framework to data from 56 industrial sectors in 43 countries between 2000 and 2014, we find that the susceptibility of individual industrial sectors to economic shocks varies greatly across countries, sectors, and time. We show that susceptibility-based predictions that take sector- and country-specific recovery into account, outperform---by far---standard econometric growth models. Our results are analytically rigorous, empirically testable, and flexible enough to address policy-relevant scenarios. We illustrate the latter by estimating the impact of recently imposed tariffs on US imports (steel and aluminum) on specific sectors across European countries. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-03-08 15:00 - 17:15 | G. Stamatescu (Graz University of Technology) | Data-driven Modelling of Large Scale Manufacturing Systems
Abstract The emergence of ubiquitous sensing and control of the physical world through networks of embedded computing devices has currently gone beyond conventional areas of ambient and environmental monitoring and reached the cost-sensitive, generally risk averse and safety critical industrial domain. Data-driven models have been deployed for process fault detection and diagnosis, improving yields of production lines and optimisation of key performance indicators such as energy efficiency and maintenance and productivity related metrics, with outreaching social and environmental impact. The talk aims to discuss the current context and challenges related to the design, implementation and evaluation of advanced data processing and learning algorithms in industry applications. Special focus is given to in situ efficient inference on distributed embedded devices by means of novel deep neural network architectures.
<p></<>
Bio Grigore Stamatescu is Associate Professor at the University Politehnica of Bucharest, Romania and a Visiting Research Scientist at the Complexity Science Hub and the Technical University of Graz. His current research interests span the areas of networked embedded sensing, the internet of things and distributed information processing in industry, the built environment and smart city applications, with more than 100 publications in international journals and conference proceedings. He was a 2015-2016 Fulbright Visiting Scholar and he is a current recipient of the Joint Excellence in Science and Humanities grant of the Austrian Academy of Sciences. Grigore Stamatescu is member of the IEEE Robotics and Automation Society, ACM and the IFAC TC 3.3. Telematics: Control via Communication Networks. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-03-01 15:00 - 17:15 | J. Preiser-Kapeller (Austrian Academy of Sciences) | Micro-histories, macro-dynamics and the contested study of medieval societies | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-01-25 15:00 - 17:15 | T. Heinrich (University of Oxford) | Agent-based models of industrial organization and technological change Abstract Technological change and industrial organization are connected on many levels: Where do new technologies emerge?, how do they diffuse in the industry and among consumers?, what impact does this have on the characteristics of the cross-sectional distribution of firms? Technologies do not evolve isolated: They rely on other technologies, on infrastructure, and on specialized knowledge. This causes network externalities and links the future success of technologies to the number of previous users of not only these technologies, but also of the pre-existing infrastructure and related technologies. Observing the emergence and diffusion of technologies is crucial for our understanding of both the role and potential of technological change and our capacity to incentivize the emergence of green technologies, encourage prudent practices in finance and insurance, and address problems such as climate change. Some results from agent-based modeling and replicator dynamics are presented and substantiated by empirical results from patenting in green technologies, as well as industrial dynamics in the insurance business and other sectors.
| CSHV, 8., Josefstädter Str. 39, room 201 |

2019-01-18 15:00 - 17:15 | T. Pham (Medical University of Vienna) | A brief review on a generalised model of epidemic spreading Abstract The aim of this talk is to revise the SWIR model which has recently been used as a general framework for various cascading dynamics, such as k-core percolation, breakdown on interdependent networks and cooperative epidemic spreading. While this model works well for static networks, the adaptive nature of many real-world processes requires to consider the network evolution as well. It will be discussed how to incorporate an adaptation into the original SWIR to uncover new phenomena that do not appear over static networks. | CSHV, 8., Josefstädter Str. 39, room 201 |

2019-01-11 15:00 - 17:15 | J. Korbel (Medical University of Vienna) | Scaling expansions: universal tool for classification of complex systems Abstract The nature of statistics, statistical mechanics and consequently the thermodynamics of stochastic systems is largely determined by how the number of states W(N) depends on the size N of the system. Here we propose a scaling expansion of the phasespace volume W(N) of a stochastic system. The corresponding expansion coefficients (exponents) define the universality class the system belongs to. Systems within the same universality class share the same statistics and thermodynamics.
For sub-exponentially growing systems such expansions have been shown to exist. By using the scaling expansion this classification can be extended to all stochastic systems, including correlated, constraint and super-exponential systems. The extensive entropy of these systems can be easily expressed in terms of these scaling exponents. Systems with super-exponential phasespace growth contain important systems, such as magnetic coins that combine combinatorial and structural statistics. We discuss other applications in the statistics of networks, aging, and cascading random walks. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-12-14 15:00 - 17:15 | M. Strauss (Complexity Science Hub Vienna) | Data-driven identification of new disease phenotypes | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-12-07 15:00 - 17:15 | L. Horstmeyer (Medical University of Vienna) | Precursor Theory of Adaptive Networks Abstract Adaptive co-evolving dynamic networks play a key role in ecological, epidemiological, social or financial systems. In adaptive network models the node states and the network topology are dynamically coupled, i.e. they co-evolve. These models have in common that they can collapse. We would like understand these collapse transitions and their precursors in adaptive network models. Structural information about the network is of crucial importance for this task. These models differ strongly with respect to the level of aggregation at which this information is required. On one side there are systems whose collapse depends solely on a very particular network configuration and on the other side there are systems whose collapse depends on aggregated structural variables, such as the overall density of certain network motifs. In this talk I contrast two representative models that differ in this respect: The Jain-Krishna model and the adaptive SIS model.
In the first model the collapse occurs when cycles are broken. We show that this collapse can be predicted accurately purely from information of the nodes, even though it relies crucially on a particular network topology. In the second model we show that a prediction may be systematically off, even in the presence of aggregated knowledge of the network topology. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-11-30 15:00 - 17:15 | B. Mutlu (KnowCenter) | Methods to Recommend and Personalize Visualizations Abstract Data analysis and data mining in particular as emerging fields of data science have gained wide popularity in the last years both in academia and in industry. As an example, recent research initiatives such as Industry 4.0, SmartFactory or Internet-of-Things (IoT) try to motivate researchers and engineers to improve the production and products in various application fields by utilizing technologies for data analysis such as clustering, filtering and visual data analysis for example. Beside many issues that have to be solved in this intent, the major problem that still remains here is how to deal with the growing amount of data/information, i.e., a so-called information overload problem. Visual data analysis has been proven to be one of the effective ways to tackle this issue. Visualizations have a distinctive advantage when dealing with the information overload problem: because they are grounded in basic visual cognition, many people understand them and can naturally perform visual operations such as clustering, filtering and comparing quantities. However, creating appropriate visualizations requires specific expertise of the domain and underlying data. Yet, an ordinary user lacks expert knowledge and can rarely generate sophisticated visualizations. Thus, the first quest of this work is to provide strategies to automatically recommend appropriate visualizations for non-experts by following visual encoding rules and perceptual guidelines. Yet, considering just visual encoding rules leads to a large set of possibilities, valid in terms of representing the data visually, but without considering which type serves the users' needs best. To tackle this issue, we propose a novel recommender system that (i) recommends visualizations based on a set of visual cognition rules and (ii) filters a subset considering the user's preferences. This research work investigates different strategies to recommend visualizations considering different aspects of the user preferences/needs/interest.
Bio: Belgin Mutlu is a senior researcher at the Know Center GmbH and area manager at the Pro2Future GmbH in Graz. She received her Master’s and Bachelor’s degree in Telematics, and her PhD degree in Informatics from Graz University of Technology. Her research interests include visualizations, visual data analysis, recommender systems and semantic web. Belgin has worked as researcher and developer in the field of adaptive visualizations, linked data and visual recommender systems in several EU Projects—CODE, EEXCESS, and AFEL. She has co-authored more than 15 peer-reviewed publications. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-11-23 15:00 - 17:15 | S. Schweighofer (Medical University of Vienna) | The Measurement of Meaning Abstract Affective disorders such as depression, anxiety disorder or PTSD are characterized by a symptom called ‘negative repetitive thinking’: Patients’ thoughts tend to circle around a small set of topics. Conversely, in a state of psychosis, patients often experience ‘flight of thought’, i.e. their thoughts leap from topic to topic, often following only very distant associations. How can we develop metrics to detect these symptoms in spoken or written communication? I discuss and compare several approaches, such as word/text embeddings models, and measures derived from large-scale word association networks. I will also give an outlook on other potential application areas, such as creativity research. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-11-16 15:00 - 17:15 | M. Miess (Vienna University of Business and Economics) | Economic Forecasting with DSGE models - an evaluation and critique Abstract New Keynesian Dynamic Stochastic General Equilibrium (NK-DSGE) models with numerous real and nominal frictions have become the standard theory-based model for economic forecasting. To explain the origins of these models in general equilibrium theory, this talk will give an overview of the theory and assumptions behind NK-DSGE models. Secondly, it will explain how these models are taken to the data using Bayesian estimation techniques, and present results of a forecast evaluation exercise with a two-country NK-DSGE model for Austria and the Euro Area. Lastly, the lecture reflects on the continued use and apparent success of NK-DSGE models in economic forecasting, despite widespread critique of their underlying theory and the way Bayesian estimation is conducted in a DSGE context. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-11-09 15:00 - 17:15 | R. Hanel (Medical University of Vienna) | Understanding driven and irreversible processes Abstract We are used to working with independent ensembles or systems exhibiting detailed balance.
If the typical system we are interested in are not of this kind then what happens with our mathematical
tools to describe such systems? And can we say anything about how irreversible a process works?
In this lecture we will briefly discuss these two questions. For the first question we briefly look at
the maximum configuration entropy of driven processes. For the second question
we will look more closely at thermodynamic cycles performed in finite time in a gas kinetic model
to find a relation between the length of cycle periods and the amount of irreversible work obey a relation
similar to a Heisenberg uncertainty relation. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-10-19 15:00 - 17:15 | F.A. Rodrigues (Universidade de Sao Paulo) | Epidemic processes in single and multilayer complex networks Abstract In this talk, we will present our last results on the modelling of rumour and disease spreading in single and multilayer networks. We will introduce a general epidemic model that encompasses the rumour and disease dynamics into a single framework. The susceptible-infected-susceptible (SIS) and susceptible-infected-recovered (SIR) models will be discussed in multilayer networks. Moreover, we will introduce a model of epidemic spreading with awareness, where the disease and information are propagated in different layers with different time scales. We will show that the time scale determines whether the information awareness is beneficial or not to the disease spreading. Finally, we will show how machine learning can be used to understand the structure and dynamics of complex networks. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-10-12 15:00 - 17:15 | S. Rehman (Vienna University of Technology) | Dependable, energy-efficient, and intelligent embedded computing for the Internet of Things Abstract In today’s smart world, the consumer products and electronics, sensors, industrial components, and other everyday objects are getting increasingly integrated and connected through the Internet backbone, and thereby enabling emerging application scenarios in the world of Internet-of-Things (IoT). The massive amount of data generated by the diverse interrelated heterogeneous devices (typically embedded systems) has to be first computed, transferred in real time via internet and take actions without requiring human or computers contact/intervention. This poses high demands on adaptability/intelligence, energy-efficiency and reliability for these systems that require advanced features to enable correct functionality and availability of the services to the user despite sudden interruptions in the system operation, device failures, harsh environments, and escalating security threats.
This talk will start with a quick overview of important dependability issues, prominent state-of-the-art techniques, and various hardware/software modeling and optimization techniques developed by me together with my research team. A key focus will be on bridging the gap between hardware and software to achieve accurate reliability models at the higher system layers while accounting for the underlying hardware features. This provides a foundation to develop and employ diverse robustness optimizations at different system layers of (heterogeneous) computing devices deployed at the Edge, under given performance and power constraints. I will also give a quick outlook of my recent works in approximate computing that allow trading output quality/accuracy with efficiency in terms of area, latency, power, and energy consumption, thereby enabling highly energy-efficient and low-latency Edge nodes. Towards the end, I will present a summary of my other professional activities and research proposals, followed by an outlook of my future vision and research agenda. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-10-05 15:00 - 17:15 | N. Lauber (University of Vienna) | Influence of the Group Size on the Rate of Combinatorial Innovation Production Abstract Innovation can often be formulated as a recombinatorial-process where objects from a set are combined to form new objects that are in turn added to the set of objects that can again be recombined to form yet new objects.
In many instances this is a social process where the recombination is performed by a group of collaborators. This is especially true when it comes to the creation of new knowledge or methods as this mostly happens by people exchanging and combining their current knowledge and ideas to create new knowledge.
For this to happen communication between the collaborators is necessary. However in social psychology one can often observe a phenomenon called Ringelmann-Effect, which states that as such groups grow in size the increasing coordination overhead reduces the rate of successful communications. This ultimately leads to less innovative exchanges and a decline in novelty production.
In this talk I want to give a short overview about a simple combinatorial innovation model that mimics the recombination of knowledge elements by a set of agents and an analysis of data from Open-Source-Software projects. The aim is to investigate the influence of the Ringelmann-Effect in the model as well as in the data-set. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-06-29 14:00 - 16:15 | J. Sorger (Complexity Science Hub Vienna) | An Introduction to Visualization Research & EuroVis 2018 Conference Report Abstract In this talk I will first explain the goals and sub domains in visualization research, followed by a short overview of some of the most compelling publications presented at this year's EuroVis 2018 conference. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-06-22 14:00 - 16:15 | S. Pfenninger (ETH Zurich) | Designing an energy system based on variable wind and solar power Abstract If average global warming is to be contained below 2°C, greenhouse gas emissions from the energy sector must be eliminated by mid-century.
Wind and solar photovoltaics provide a large untapped resource and have experienced rapid cost reductions.
They are thus poised to form the backbone of a clean energy system. However, questions remain about how to integrate them into a stable, affordable and sustainable energy system at scales ranging from cities to continents, in particular due to their weather-dependency, the resulting requirements to spread them spatially, and the additional technologies like electricity storage needed for their integration.
This presentation will provide an overview of ongoing work to model these technologies and their integration into the energy system, and the key open challenges to successfully manage the renewable energy transition. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-06-15 14:00 - 16:15 | J. Neidhardt (Vienna University of Technology) | Social Influence and Opinion Forming in Online Social Networks Abstract Web-based phenomena such as fake news, hate speech or echo chambers have been gaining increasing attention in the public discourse. However, the extent and actual impact of these phenomena are still unknown. In order to address these issues in more depth, it is important to understand how opinions emerge, change and get exchanged. In our work, we therefore intend to develop novel, data-driven approaches to examine social influence mechanisms and opinion forming at a large scale. We aim, in particular, to introduce a complex user model, which characterizes each person with respect to three levels: 1) the individual level capturing the characteristics of a user; 2) the network level capturing the social relationships of a user; and 3) the group level describing sets of similar users with certain dominant opinions and high-level ideologies. For the empirical analysis we use a rich and unique dataset of an online news forum that contains the complete posting history of all users, their detailed behaviors and interactions as well as all news articles and the respective meta data for almost two decades. Short Bio: Julia Neidhardt is a researcher at the division of E-Commerce at TU Wien. She holds a Master's degree in mathematics from the University of Vienna and a PhD in Computer Science from TU Wien. Her research focuses on modeling and predicting complex human behavior, user preferences and social relations as well as their dynamics in digital-enabled environments. Current projects are concerned with social influence in online communities, the emergence and diffusion of topics, opinions and sentiments, picture-based travel recommender systems, and social media-based event prediction. Since 2013, she has been regularly visiting the Science of Networks in Communities (SONIC) research group at Northwestern University, USA. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-06-01 14:00 - 16:15 | N. Haug (Medical University of Vienna) | Forecasting human disease trajectories with healthcare networks Abstract The availability of large scale healthcare data promises to revolutionise medical care towards a precision medicine where treatments are tailored individually to each patient. In the last decade, the analysis of medical records, often recorded for billing purposes, was centred around binary relations between diseases, such as the comorbidity between them. The concept of comorbidity quantifies the probability that a patient with disease A also has disease B, but does not capture the long term history of a patient. For example, the subsequent development of diseases A,B and C could indicate the presence of a particular genetic defect, increasing the likelihood for disease D. The aim
of our work is to bridge this gap by developing a method to forecast
future diagnoses of a patient which takes into account the entire history, i.e., the specific sequence, of his or her diagnoses. To this end, we analyse a data set containing for each inpatient hospital stay of patients in Austria over two years the main and side
diagnoses and the admission and release dates. The data thus contains a trajectory on the set of diseases for each of the approximately 8M people insured via an Austrian social security
provider. We represent the health history of each patient by a binary vector, and divide the set of patients into categories by using a hierarchical clustering algorithm. When patients acquire new diseases, they can change their cluster. We analyse the pattern of transitions between the different clusters, in particular, we see that there exists a stream of patients into clusters acting as sinks, characterised by highly multimorbid patients. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-05-25 14:00 - 16:15 | P. Klimek (Medical University of Vienna) | Counter-dominance signaling drives evolution of cultural elites: quantitative evidence from fashion cycles in music Abstract Human symbol systems such as art and fashion styles emerge from complex social processes that govern the continuous re-organization of modern societies. They provide a signaling scheme that allows members of an elite to distinguish themselves from the rest of society. Efforts to understand
the dynamics of art and fashion cycles have so far been based on costly signaling theory, where elite members signal their superior status by introducing new symbols (e.g. fashion-styles), which are subsequently adopted by low-status groups. In response to this adoption, the elite members need to introduce yet new symbols to signal their status. We propose an alternative explanation based
on counter-dominance-signaling. There, members of the elite want others to imitate their symbols; changes only occur when outsider groups successfully challenge the elite by introducing signals that contrast those endorsed by members of the elite. To clarify the mechanism that actually drives fashion cycles in musical styles, we use a dynamic network approach on data containing almost 8 million musical albums released between 1956 and 2015. There a network systematically quantifies artistic similarities of competing musical styles. By studying the dynamics of the network we can formulate statistical hypothesis tests for whether new symbols are introduced (i) by current elite members as predicted by costly signaling theory or (ii) as a consequence of challenges by peripheral groups through countersignals. We find clear evidence that counter-dominance-signaling drives changes in musical styles. This provides a quantitative, completely data-driven answer to a century old debate about the nature of the underlying social dynamics of art and fashion cycles. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-05-18 14:00 - 16:15 | D. García (Medical University of Vienna) | Affective Intelligence in Online Political Discussion Forums
Abstract Affective Intelligence Theory (AIT) postulates that political behavior is strongly influenced by affective states.
Enthusiasm and aversion lead people to rely on simple cognitive heuristics and reinforce social identity, whereas anxiety causes people to adopt more complex cognitive strategies and weakens group boundaries.
Until now, AIT has mostly been tested on survey data.
Instead, we tested a dimensional reformulation of AIT (DAIT) against data from a German-speaking online news community, 20min.ch.
In this community, users discuss recent news and give up and down votes to each others' comments, often producing heated discussions.
We find that discussions characterized by low affective potency and extreme valence ("enthusiasm" and "aversion") produce more polarized votes and signal lower cognitive complexity than discussions with high potency ("anxiety") and/or less extreme valence.
This effect is even stronger when there is a high salience of social identity in discussions.
Thus, we can confirm DAIT, and open the door for further analysis of DAIT in other languages and settings. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-05-04 14:00 - 16:15 | J. Korbel (Medical University of Vienna) | 1) Consistency of the Maximum Entropy Principle for Generalized Entropies; 2) Transitions between superstatistical regimes Abstract 1) I will discuss the consistency of Maximum entropy principle for generalized entropies (e.g. Tsallis or Rényi entropy). Recently, several authors discussed the statistical consistency of general entropies with different results. During the talk, we show that the class of entropies obeying the consistency axioms is relatively large and enables us to model systems with inherent correlations.
2) Superstatistics is a tool which enables us to model processes on different scales assuming the existence of local equilibria. In the original definition are only two scales present. We show how to generalized the superstatistics to arbitrary number of characteristics scales, leading into transition of superstatistics for different (time)-scales. We also demonstrate the transition on the example of high-frequency data from financial markets. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-04-27 14:00 - 16:15 | C. Molinero (Complexity Science Hub Vienna, Austrian Institute of Technology) | The efficiency of road networks Abstract Measuring the efficiency of a road network is currently an open question and many approaches have been suggested. In this paper we tackle this issue by proposing a network generation model, studying its properties and laying out a definition for the performance of a network. We then confront the real road network against the highest performance version of this model and define efficiency as the ratio between the performance of the real road network and our idealised model.
This is obtained through the definition of local and weighted versions of centrality measures. These measures deal with distance-decay effects and nodes having different masses.
In doing so, we remark that each possible network created with our model has a typical fractal dimension, and this permits to map a fractal dimension (a way to occupy space) with a performance. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-04-20 14:00 - 16:15 | L. Horstmeyer (Medical University of Vienna) | Novel insights into the precursor theory of adaptive networks Abstract The adaptive SIS model is one of the most elementary models in which the
dynamics on the network is coupled to the dynamics of the network. We
investigate the critical behavior of the network topology by looking at
the densities of motives, the clustering, the compactness, the degree
distribution and assortativity, the effective branching ratio and spectral distribution. The proximity to the persistence threshold can be sensed in some of these quantities long before the transition via the appearance of local maxima due to a crossover phenomenon. This is also reconcilable with the pair approximation. A good signal-to-noise ratio makes them novel candidates for precursors of the transition. We explain
the mechanism that gives rise to this behavior and some caveats. We also
discuss the issues that arise when fitting power laws to these quantities. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-04-13 14:00 - 16:15 | M. Paluš (Academy of Sciences of the Czech Republic) | Information transfer across time scales Abstract “More is different,” wrote P.W. Anderson [1] in order to characterize the behaviour of complex systems, consisting of many interacting elements, which cannot be explained by a simple extrapolation of the laws describing the behaviour of a few elements. In order to understand complex dynamics and emergent phenomena we need to understand interactions among system components. Interacting subsystems can mutually exchange information and influence each other; or one system can causally influence another one by a directed information flow. The mathematical formulation of causality in measurable terms of predictability was given by the father of cybernetics N. Wiener [2] and formulated for time series by C.W.J. Granger [3]. The Granger causality is based on the evaluation of predictability in bivariate autoregressive models. This concept has been generalized for nonlinear systems using methods rooted in information theory [4, 5]. Complexity observed in such systems as the human brain or the Earth climate stems not only from the fact that they consist of many subsystems. Their variability covers large ranges of spatial and temporal scales and the nonlinear character of these systems leads to interactions of dynamics across scales. For instance, dynamical processes on large time scales influence variability on shorter time scales. In order to detect cross-scale causal interactions we have recently introduced a methodology [6] which starts with a wavelet decomposition of a multi-scale signal into quasi-oscillatory modes of a limited bandwidth, described using their instantaneous phases and amplitudes. Then their statistical associations are tested using the information-theoretic formulation of the Granger causality. The analysis of long-term air temperature records uncovers causal influence and information transfer from large-scale modes of climate variability with characteristic time scales from years to almost a decade to regional temperature variability on short time scales. The phenomenon of cross-scale interactions has non-negligible influence on the air temperature variability in the European mid-latitudes [7]. In the tropical Pacific, the interactions of the annual cycle with slower modes of atmospheric and oceanic variability are of special interest in order to better understand the El Niño Southern Oscillation. The cross-frequency interactions and information transfer can be observed in brain dynamics, where the cross-frequency coupling enriches the cooperative behaviour of neuronal networks and apparently plays an important functional role in neuronal computation, communication, and learning.
[1] P. W. Anderson, Science 177, (1972) 393
[2] N. Wiener, in: E. F. Beckenbach (Editor), Modern Mathematics for Engineers (McGraw-Hill, New York, 1956)
[3] C.W.J. Granger, Econometrica 37 (1969) 424
[4] K. Hlavá?ková-Schindler et al., Phys. Rep. 441 (2007) 1
[5] M. Paluš, M. Vejmelka, Phys. Rev. E 75 (2007) 056211
[6] M. Paluš, Phys. Rev. Lett. 112 (2014) 078702
[7] N. Jajcay, J. Hlinka, S. Kravtsov, A. A. Tsonis, M. Paluš, Geophys. Res. Lett. 43(2) (2016) 902–909 | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-03-23 14:00 - 16:15 | R. Boyd (University of Texas, Austin) | The Arc of Narrative: The Objective Quantification of Story Structure Abstract Scholars have long debated the existence and properties of a “common structure” that underlies narratives. Using modern, computer-based language analysis methods, we measured several structural and psychological categories of language across multiple corpora of novels, short stories, films, and written projective stories. Across all forms of stories, analyses revealed a small but consistent underlying structure to narratives that is driven by 3 primary processes: Staging, Plot Progression, and Cognitive Tension. This research provides empirical insights into the structure of narrative, providing hints as to what it may tell us about the evolution of story-making. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-03-16 14:00 - 16:15 | S. Schweighofer (ETH Zurich) | Affect and Political Polarizaton Abstract Polarization poses an existential threat to democratic political systems. Until recently, polarization was analyzed purely in terms of political positions. However, it becomes more and more apparent that polarization must also be seen as an affective phenomenon. In this presentation, we show that affect 1) correlates with polarization on the macro-level of Swiss political history, quantified with a novel measure of relational polarization, 2) increases the repetitiveness of online discussions, quantified on the basis of a neuro-probabilisitic language model, and 3), facilitates the alignment of issue positions. We also present an agent based model that explains the coupling between affect and opinion alignment, and produces testable psychological hypotheses. Our outcomes suggest that affect is not only an epiphenomenon, but a causal driver of political polarization. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-03-09 14:00 - 16:15 | S. Monica (University of Parma) | KTMAS: Analytic Modelling of the Dynamics of Multi-agent Systems Abstract Multi-agents systems (MAS) have been introduced in the field of artificial intelligence (AI) to denote a set of entities which act and interact in intelligent (i.e., often: rational) ways. Beyond AI, the application of results concerning MAS involves various other scientific disciplines, such as: biology; sociology; finance; and
economics.
In the literature, a large number of models have been proposed to describe the dynamics of MAS. Most of these are based on simulations, limiting their validity to the particular scenarios
considered and to the specific choice of simulation parameters: In order to derive more robust results, it is of interest to identify viable analytic approaches.
Our framework is inspired by mathematical kinetic theories,
with the aim to obtain analytic results on the dynamics of MAS from the description of the effects of single interactions among agents. KTMAS can incorporate different types of interactions, and it can be used to describe the dynamics of different features of MAS.
We present our research in the scope of the new disciplines of econophysics, which can be used to model wealth evolution, and sociophysics, which can be adopted to characterise opinion evolution. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-01-26 14:00 - 15:30 | P.M. Ettel (Medical University of Vienna) | tba | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-01-19 14:00 - 15:30 | V. Palchykov (National Academy of Sciences of Ukraine) | Human mobility and communication patterns: agent-based modeling approach Abstract In this talk I will present a fresh overview of my past projects on human behavior with the emphasis on agent-based modeling. Among the topics I will focus on the project on inferring human mobility using communication records. Our results show that human mobility is well predicted by a simple model based on the frequency of mobile phone calls between two locations and their geographical distance. We argue that the strength of the model comes from directly incorporating the social dimension of mobility. Furthermore, as only aggregated call data is required, the model helps to avoid potential privacy problems. | CSHV, 8., Josefstädter Str. 39, room 201 |

2018-01-12 14:00 - 15:30 | N. Haug (Medical University of Vienna) | Predicting human disease trajectories Abstract The availability of large scale healthcare data, promises to revolutionise medical care, with treatments tailored individually to each patient. In the last decade, the analysis of medical records, often recorded for billing purposes, was centered around binary relations, such as the comorbidity, between two diseases. Comorbidity quantifies the probability that a patient with disease A also has disease B. While this measure is often sufficient for practical purposes, in other cases it may be beneficial to consider the long term history of a patient, that is the interplay of more than two diseases, and their temporal order. For example, the simultaneous presence of diseases A,B and C could indicate the presence of a particular genetic defect, increasing the likelihood for disease D. To our knowledge, the time sensitive analysis of interactions of more than two diseases has not yet been pursued systematically in the literature. The aim of our work is to bridge this gap by developing a method to predict future diagnoses of a patient which takes into account the entire history of the patient's diagnoses. To this end, we analyse a data set containing for each inpatient hospital stay in Austria the ID of the patient, the main and side diagnoses and the admission and release dates for the years 2006 and 2007, and the same information for Lower Austria for the years between 2008 and 2011. The data thus contains a "disease trajectory" for each of the around 8M people insured via an Austrian social security provider. This data can be combined with data for health care provider (e.g. GP, dermatologist, ...) visits and data on medical prescriptions, which is available to us for the same patient cohort. In this talk I will explain our current ideas on how to tackle the problem of diagnosis prediction beyond pairwise statistics. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-12-01 14:00 - 15:30 | L. Horstmeyer (Medical University of Vienna) | On the Jain and Krishna model Abstract In this talk we will discuss the catalytic network model that Jain and Krishna introduced in 1998 as a toy-model of evolution. This model exhibits cycles of growth, organisation and collapse. After pointing out that it is much more generic than expected, we will exhibit some of the interesting questions that pop out of its rich structure: Can one anticipate the length of the organised phase or predict the time and strength of a collapse? To this end we will make a short excursion to the statistics of binary threshold tests and then tackle some questions regarding the prediction of catastrophic events in this model. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-11-24 14:00 - 15:30 | N. Lauber (University of Vienna) | Do small evolutionary systems innovate faster? Abstract A recent path to understand the dynamics of evolutionary systems is “combinatorial evolution". The essence of this approach is, that an evolving system can be formulated as a set of objects, which can be combined to pairs to form new objects that are in turn added to the set and can be again combined to form yet new objects and so forth. In this talk I want to give a short overview about a simple combinatorial evolution model how knowledge in the form of novel ideas is created through the recombination of preexisting ideas. I especially want to focus on how the performance of the system depends on its size and draw parallels to how the efficiency of groups in producing new knowledge might depend on the size of the group. | CSHV, 8., Josefstädter Str. 39, room 255 |

2017-11-17 14:00 - 15:30 | O. Saukh (Complexity Science Hub Vienna, Graz University of Technology) | Sensing the Air We Breathe Abstract Air pollution is traditionally monitored by networks of static measurement stations operated by official authorities. These stations are highly reliable and accurately measure a wide range of air pollutants with expensive analytical instruments. Large size, high price, and laborious maintenance of these complex measurement systems severely limit the number of installations resulting in a low spatial resolution of the available pollution information. In the last few years, low-cost solid-state gas sensors became available on the market. We integrated these sensors in ten lightweight air quality monitoring stations and deployed them on top of trams in Zurich to increase spatial coverage of the urban area. In this talk I will focus on three main challenges we faced in this deployment: 1) pre-deployment sensor testing and compensation for cross-sensitivities, 2) automatic calibration of low-cost sensors to improve data quality, and 3) data analysis and predictions to construct high resolution air pollution maps. I will give an overview of the theory behind the developed methods and evaluate their performance on a data set comprising 400 million measurements collected in Zurich over three years. Generated pollution maps received a lot of attention in the local media and are now part of the public service offered by the City of Zurich. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-11-10 14:15 - 15:45 | S. Poledna (Institute for Applied Systems Analysis - IIASA) | Linking natural disasters with socio-economic systemic risk
through a data-driven agent-based model Abstract This paper describes a novel approach for estimating the indirect economic losses due to natural disasters with an application to flood events in Austria. The approach combines a probabilistic physical damage catastrophe model with a macroeconomic agent-based model (ABM). The ABM methodology advances state-of-the-art approaches by exploiting large data sets (big data) with close to a million agents (households, non-financial and financial firms and a general government) calibrated with data from national accounts, input-output tables, government statistics, census data and business information. The probabilistic flood model is equally innovative by introducing a copula methodology that provides an assessment of flood losses by taking account of spatial dependencies in the flood hazard. The ABM includes an input-output model with 64 industries where all goods and services are produced endogenously, and the probabilistic copula approach provides a nation-scale estimate of direct flood losses over the full risk spectrum based on basin-scale loss distributions and exposure with Corine land-cover mapping. The direct loss estimates are used to build a damage scenario generator that provides the input for the ABM, which, in turn, assesses the indirect economic losses due to the event. The analysis shows that there can be severe indirect economic losses in Austria due to large-scale natural disasters, or systemic events, and shows the impact chains leading to the systemic losses. One of the main findings is that the distinct types of disaster events exhibit qualitatively different economic behavior: while more moderate scenarios induce positive indirect economic effects in the medium term, the severe, or systemic, events result predominantly in a negative economic response throughout the simulation period. Most importantly, the analysis disaggragates the gains and losses occurring to different sectors. This detailed information can be useful for assessing risk management options at various scales. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-11-03 14:00 - 15:30 | D. García (Complexity Science Hub Vienna, Medical University of Vienna) | Understanding Collective Emotions through Digital Traces of Human Behavior Abstract Collective emotions are emotional states temporarily shared by large amounts of individuals. From riots to sport events, from viral content to online quarrels, collective emotions are subject to appear in various situations and have long-lasting consequences. Understanding collective emotions has been challenging due to their fast evolution, large scale, and complex dynamics, limiting their tractability in natural scenarios and their controllability in experiments. Recent developments in computational social science, in particular agent-based modeling and sentiment analysis of digital traces, allow us to quantify and model emotions at unprecedented scales and resolutions, offering new opportunities to study collective emotions.
I will present an overview of modeling and analysis of collective emotions at scale, with an emphasis on how these can be observed through social media and how they can be reproduced in computational models. I will illustrate the application of digital traces of collective emotions to understand the collective responses to the Paris terrorist attacks of November, 2015. Using a large-scale dataset of Twitter public messages, we measured affect as expressed through text. We find the traces of the simultaneous negative reaction to the attacks, which are later replaced by a synchronization process that increases the salience of shared values and prosocial behavior. Our analysis supports the hypothesis that synchronization leads to social resilience after a collective trauma, illustrating the social function of collective emotions. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-10-27 14:00 - 15:30 | L.J. Uberti (University of Oslo) | The Corruption U-Curve Abstract We replicate Saha and Gounder (Econ Model 31: 70-79, 2013), who show that the relationship between corruption and economic development exhibits an inverted U-shaped pattern. Pooled OLS estimates, however, should not be interpreted as conclusive evidence in support of a causal relationship. Using a much longer panel of 156 countries during 1900-2010, we show that a U shape shows up even after controlling for country-level heterogeneity, suggesting that the estimated relationship does not result from a version of the classic “Kuznets fallacy”. The curvilinear relationship is also robust to controlling for potential simultaneity using an instrumental variable approach. Still, the parabola implied by a fixed-effects model is considerably less “peaked” than the one implied by the corresponding OLS model.
| CSHV, 8., Josefstädter Str. 39, room 201 |

2017-10-13 14:00 - 15:30 | J. Korbel (Medical University of Vienna) | Information flows between communities in complex financial networks Abstract With the help of transfer entropy, we analyze information flows between communities of complex financial networks. We show that the transfer entropy provides a coherent description of interactions between communities, including non-linear interactions. We analyze transfer entropies between communities of five largest financial markets, represented as networks of interacting stocks. Additionally, we discuss information transfer of rare events, which can be analyzed by Rényi transfer entropy. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-10-06 14:00 - 15:30 | K. Kaski (Aalto University School of Science, Finland; Wolfson College, Oxford University, UK; Complexity Science Hub Vienna, Austria) | Circadian rhythms of urban people – are we like fruit flies? Abstract All living organisms, including humans, have internal biological or circadian clock that helps them anticipate and adapt to the regular rhythm of the day. The timings of human activities are marked by circadian clocks which in turn are entrained to different environmental signals. In an urban environment, the presence of artificial lighting and various social cues tend to disrupt the natural entrainment with the sunlight. However, it is not completely understood to what extent this is the case. Here we exploit the large-scale data analysis techniques to study the mobile phone calling activity of people in large cities to infer the dynamics of urban daily rhythms. From the calling patterns of about 1,000,000 users spread over different cities but lying inside the same time-zone, we show that the onset and termination of the calling activity synchronizes with the east-west progression of the sun. We also find that the onset and termination of the calling activity of users follows yearly dynamics, varying across seasons, and that its timings are entrained to solar midnight. Furthermore, we show that the average mid-sleep time of people living in urban areas depends on the age and gender of each cohort, most likely as a result of biological and social factors. | CSHV, 8., Josefstädter Str. 39, room 201 |

2017-06-30 14:15 - 15:45 | A. Hinteregger (University of Vienna) | Systemic risk in the Austrian economy Abstract Systemic risk is the notion that a shock may not only render directly affected components but (possibly large) parts of the network non-functioning due to cascading effects. In a financial network this systemic risk stems from the exposures of lending institutes to one another. Network measures such as DebtRank were used to estimate the possible impact on the interbank network resulting from a market shock. In this work, I extend the analysis of systemic risk from financial institutes to the whole economy. and find that the DebtRank distribution of companies is qualitatively similar to the distribution of financial institutes. Companies may have exposures to a disadvantageous set of banks, thus enabling the propagation of an impact through large parts of the network. Even though larger companies tend to have a higher DebtRank, there are companies sharing a very similar DebtRank with total assets that span multiple orders of magnitude. The results suggest that from a systemic risk perspective companies and banks have some similarities. It may therefore be beneficial to extend regulations that try to alleviate systemic collapses to the whole economy. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-06-23 14:15 - 15:45 | S. Poledna (International Institute for Applied Systems Analysis - IIASA) | Economic forecasting with an agent-based model Abstract We develop an agent-based model (ABM) for the Austrian economy using data from national accounts, input-output tables, government statistics, census data and business surveys. The model incorporates all economic activities (producing and distributive transactions) as classified by the European system of accounts (ESA) and all economic entities, i.e. all juridical and natural persons, are represented by agents (at a scale of 1:10). We show that this model is able to compete with vector autoregressive (VAR) and autoregressive–moving-average (ARMA) models in out-of-sample prediction. Potential applications of this ABM include economic forecasting, as well as the prediction of responses of the economy to endogenous shocks, e.g. from the financial system, or exogenous shocks like natural disasters, transformative technological innovations or unintended consequences of political interventions such as subsidies and tax policies. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-06-16 14:15 - 15:45 | A. Pichler (Vienna University of Economics and Business) | Minimization Systemic Risk as an Optimal Network Reorganization Problem - The Case of Overlapping Portfolio Networks in the European Government Bond Market Abstract Systemic risk arises as a multi-layer network phenomenon, where layers represent direct financial exposures of various types, including interbank liabilities, derivative- or foreign exchange exposures. Another network layer of systemic risk emerges through common asset holdings of financial institutions. Strongly overlapping portfolios lead to similar exposures caused by price movements of financial assets. We use a simple method to quantify systemic risk of overlapping portfolio networks from endogenous asset sales within the network. We then present a general optimization procedure where we minimize the systemic risk in a given financial market by optimally re-ordering overlapping portfolio networks, under the constraint that the expected returns and risks of the individual portfolios is unchanged. We explicitly demonstrate the method on the overlapping portfolio network of sovereign exposure between major European banks by using data from the European Banking Authority stress test 2016. We show that multiple systemic-risk-efficient allocations do exist, which are actually accessible by the optimization and that do not alter the returns and risks of the institutions’ portfolios. In the case of sovereign exposure, systemic risk can be reduced by more than 56%, without any detrimental effects. We confirm these results by a simple simulation of fire-sales in the government bond market. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-06-09 14:15 - 15:45 | L. Horstmeyer (Medical University of Vienna) | The Gillespie Algorithm Abstract How can one numerically simulate large stochastic networks? Perhaps
one would go through each node in the network and update its state
with a probability dictated by the denition of the model. This may
not always be very ecient and can also lead to biases. The Gillespie
algorithm provides a great alternative. I will present its merits and its
implementation. The talk will close with the illustration of the Gillespie
algorithm for an adaptive stochastic network dynamics, namely the SIS
model with rewiring. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-05-26 14:15 - 15:45 | C. Tsallis (Centro Brasileiro de Pesquisas Físicas) | Nonadditive entropy: small price to satisfy thermodynamics -- Theory and experiments Abstract The Galilean composition law of velocities within Newtonian mechanics is additive. But, in order to unify mechanics with Maxwell electromagne+sm, Einstein adopted the Lorentz space-time transformation as the primary mathematical-physical goal to be satisfied. It resulted the well known, nonadditive, relativistic composition law of velocities. This surely is a small price to pay in order to unify mechanics and electromagnetism, and explain very many experimental facts. Analogously, there is a plethora of analytical, experimental, observational and computational evidences (see Bibliography in h]p://tsallis.cat.cbpf.br/biblio.htm) which reveals various kinds of violations of Boltzmann-Gibbs statistical mechanics, including thermodynamics. The adoption of nonadditive entropic functionals which generalize the traditional, additive, Boltzmann-Gibbs entropy enables to satisfy classical thermodynamics: as before, a small price to pay! | CSHV, 8., Josefstädter Str. 39, Salon |

2017-05-19 14:15 - 15:45 | E. Valdano (Universitat Rovira i Virgili) | The spread of diseases on time-evolving networks:
from models to data and back Abstract A wide range of physical, social and biological phenomena can be expressed in terms of spreading processes on networked systems. Notable examples include the spread of infectious diseases through direct contacts, the spatial propagation of epidemics driven by mobility networks, the spread of cyber worms along computer connections, or the diffusion of ideas mediated by social interactions. All these phenomena arise from a complex interplay between the spreading process and the network’s underlying topology and dynamics. Understanding how the time-evolving properties of the network impact the spread of the disease is a crucial step to setting up control and prevention strategies. The increasing availability of highly-resolved interaction data has made it possible to target a wide variety of settings and diseases, but at the same time new methodological challenges have arisen. In particular, a fundamental property of such phenomena is the presence of an epidemic threshold, i.e., a critical transmission probability above which large-scale propagation occurs, as opposed to quick extinction of the epidemic-like process. Computing this threshold is of utmost importance for epidemic containment and control of information diffusion. I will present a new analytical framework for the computation of the epidemic threshold for an arbitrary time-varying network. By reinterpreting the tensor formalism of multi-layer networks, this framework allows the analytical calculation of the epidemic threshold, without making any assumption on contact structure and evolution, and can be applied to a wide class of diseases.
Along these theoretical developments, challenges related to the analysis and elaboration of the increased amount of data have emerged. I will present the case of diseases affecting farmed cattle. Such diseases compromise both human and animal health and welfare, and represent a major cause of loss in economic revenue. As a result, studying the networks of animal movements is a key step in devising new prevention and containment strategies. Past works have already analyzed cattle networks in several European countries. A comprehensive study, showing the impact of country-specific driving factors on network evolution and topology, is however still missing. I will present a collaborative platform for analyzing and comparing networks from several European countries. Using a bring code to the data approach, our platform overcomes the strict regulations preventing data sharing, and allows an effective comparative analysis. I will describe the framework, and present the result of this analysis, highlighting both properties that are characteristic of livestock markets, and country-specific features. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-05-12 14:00 - 15:30 | T. Biro (WIGNER Research Centre for Physics) | Unidirectional and Resetting Stochastic Dynamics Abstract A particular class of linear master equation dynamics with unidirectional internal evolution rates and a long-jump resetting to the zero-state will be discussed. This model class is able to describe several distributions observed in complex systems, derived from the particular state dependence of the transition rates. We select out the exponential (geometrical) and the Waring distribution, along with their large system - continuous model counterparts, the Boltzmann-Gibbs and the Pareto distribution, as examples, respectively. A list of suitable properties for defining an entropic distance measure will also be presented. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-05-05 14:15 - 15:45 | L. Zavojanni (Medical University of Vienna) | Dynamical Stationarity in Processes with Sustained Random Growth Abstract In sustained growth with random dynamics, stationary distributions can exist without detailed balance. This suggests thermodynamical behaviour in fast growing complex systems. In order to model such phenomena, it is possible to use both a discrete and a continuous master equation. The derivation of
elementary rates from known stationary distributions is a generalisation of the fluctuation–dissipation theorem. Entropic distance evolution is given for such systems. Depending on the transition and loss rates, the distributions for growing networks,
particle production, scientific citations and income distribution
can be reconstructed. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-04-28 14:15 - 15:45 | R. Hanel (Medical University of Vienna) | Process, history dependence, and entropy Abstract In different fields entropy has been conceptualized in different ways resulting in a functional expression equivalent to Shannon-entropy. Therefore entropy often is thought of as a universal concept. This however is true only as long as the underlying processes in question are essenially Bernoulli processes leading to a degeneracy of most commonly used entropy concepts and the ubiquitous emmergence of the particular functional form of Shannon-entropy. If the underlying process class in question becomes history dependent, also the expressions for the entropy functionals change and it becomes important which concepts we are using in order to define a notion of entropy; whether we think of an extensive property of matter, information rates, or maximum configuration principles, for instance, then starts to make a difference. For history dependent processes entropy is no longer uniquely defined and different entropy concepts have to be distinguished, informing us about distinct properties of the process. We discuss the general situation with simple history dependent example processes. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-04-07 14:15 - 15:45 | J. Kertesz (Central European University Budapest, external fellow of the Complexity Science Hub Vienna) | Multiplex Modeling of the Society
Abstract The society has a multi-layered structure, where the layers represent the different contexts resulting in a community structure with strong overlaps. To model this structure we begin with a single-layer weighted social network (WSN) model showing the Granovetterian correlations between link strength and topology. We find that when merging such WSN models, a sufficient amount of inter-layer correlation is needed to maintain these correlations, but they destroy the enhancement in the community overlap due to multiple layers. To resolve this, we devise a geographic multi-layer WSN model, where the indirect inter-layer correlations due to the geographic constraints of individuals enhance the overlaps between the communities and, at the same time, the Granovetterian structure is preserved.
The network of social interactions can be considered as a multiplex from another point of view too: each layer corresponds to one communication channel and the aggregate of all them constitutes the entire social network. However, usually one has information only about one of the channels, which should be considered as a sample of the whole. We show by simulations and analytical methods that this sampling may lead to bias. For example, while it is expected that the degree distribution of the whole social network has a maximum at a value larger than one, we get with reasonable assumptions about the sampling process a monotonously decreasing distribution as observed in empirical studies of single channel data. We analyse the far-reaching consequences of our findings. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-03-31 14:15 - 15:45 | P. Klimek (Medical University of Vienna) | Economic resilience quantified Abstract Economic and financial crises are extremely costly, yet our understanding of the resilience of economies is only in its infancy. Resilience is typically understood as a system’s ability to absorb, withstand, and also recover from adverse events. The characterization of economic resilience is particularly challenging due to strong interdependences between different industries and the way they absorb and recover from production shocks. In this talk we develop for the first time a quantitative, predictive, and data-driven formalism for the resilience of economies based on linear response theory (LRT). Key to this formalism is Leontief’s input-output model that depicts inter-industry relationships within an economy. We show how the impacts of production shocks in specific industries can be analytically derived from small-scale fluctuations observed in the input-output model using LRT. For each industry sector in 43 different countries, we can compute its time-dependent response to different kinds of shocks in any other sector. That is, a central result of this framework is the derivation of resilience curves for each sector with respect to shocks in any other sector. The so-derived resilience curves are shown to be predictive for, both, the average growth of and fluctuations in the output in a given country. Furthermore, we show that the impact of sector-specific demand shocks on the entire economy can also be understood and to some extent predicted from these curves. The predictions are shown to be particularly accurate when the framework is applied to the financial crisis in 2008. Our work establishes a firm and novel link between the out-of-equilibrium behavior of severely disrupted economies and their steady-state fluctuations. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-03-24 14:15 - 15:45 | B. Corominas-Murtra (Medical University of Vienna) | Characterization and statistical footprints of Open-Ended Evolution
Abstract A major problem for evolutionary theory is understanding the so called open-ended nature of evolutionary change, from its definition to its origins and consequences. Open-ended evolution (OEE) refers to the unbounded increase in complexity that seems to characterise evolution on multiple scales. In this talk we will present a fundamental characterisation of OEE. Essentially, it is assumed intrinsic unpredictability and the need for an always increasing amount of information to explain the successive evolutionary steps —the emergence of innovation. Interestingly, such unpredictability defines the boundary conditions for a mathematical problem which ends with a prediction: The statistical counterpart of the OEE ‘postulates’, based on standard Shannon Information theory, have the structure of a variational problem which is shown to lead to Zipf’s law as the expected consequence of an evolutionary processes displaying OEE. Interestingly, many complex systems candidates of displaying OEE, from language to proteins, share this common scaling behaviour. Other information-theoretic phenomena arising from open-ednedness, such as the paradox of information loss, will be also discussed. We will finish discussing the connection of this general framework with existing models for the understanding of the emergence of innovation. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-03-17 14:15 - 15:45 | V. Servedio (Complexity Science Hub Vienna) | Evidence of Weber-Fechner law in political opinions
Abstract The widespread use of internet allows to conduct experiments
in the frame of social science with the engagement of a large
number of participants. In this presentation I will show the
results of a web-experiment intended to uncover voters'
perception of the Italian political space in the early 2013,
right before last Italian political elections. We find that
political opinions follow a sort of Weber-Fechner law, known
to occur in the perception of our five senses, i.e., our
perceptions depend on the logarithm of stimuli intensity.
Such finding can be exploited in the future to devise more
realistic agent-based models of opinion dynamics. | CSHV, 8., Josefstädter Str. 39, room 101 |

2017-01-27 14:15 - 15:45 | R. Hanel (Medical University of Vienna) | Extortion strategies in iterated prisoners dilemma Abstract A brief introduction to the concept of extortion strategies in iterated prisoners dilemma games and recent insights into the evolution of extortion and cooperation. Lit: (i) WH Press & FJ Dyson PNAS 109 (2012) and (ii) C Hilbe, MA Nowak & K Sigmund PNAS 110 (2013)
| MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2017-01-20 14:15 - 15:45 | A. Hinteregger (Medical University of Vienna) | Cancelled! Systemic risk in the Austrian economy | MUW, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor, room 513 |

2016-12-16 14:15 - 15:45 | M. Sadilek (Medical University of Vienna) | Identifying the driving processes of coupled friendship and enmity dynamics in a two-layer network model Abstract With the advent of social media it has become possible to study human social relations in a quantitative way. However, in most cases only data on positive relations (like friendship) are available while social balance theory states that in a social network positive and negative relations strongly depend on each other.
In the massive multiplayer online game PARDUS players can mark each other not only as friends but also as enemies, leading to a two-layer multiplex network structure.
We discuss the dynamics of friendship and enmity relations between thousands of players in PARDUS. We identify and quantify the driving processes of the associated two-layer social network formation. Well known sociological hypotheses like ‘The enemy of my enemy is my friend’ turn out to be important building blocks of understanding the dynamics of the coupled formation of friendly and hostile interactions within a society. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-11-25 14:15 - 15:45 | B. Baumann (Medical University of Vienna) | Optical imaging applications in the eye and brain Abstract Optical imaging enables rapid in vivo imaging with micrometer scale resolution. In this seminar, I will present some recent technological advances and optical imaging applications in the eye and brain. In particular, we will focus on functional extensions of a technique called optical coherence tomography (OCT). OCT enables imaging tissue microstructure as well as polarization properties and perfusion in real time. Several applications for quantitative imaging of pathological changes in humans and small animals will be presented. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-11-18 14:15 - 15:45 | S. Kitzler (Vienna University of Technology) | Comparison of methods to calculate diffusion in complex media Abstract Content of my project work and the seminar presentation is to ascertain the probability of the presence of particles which move with Brownian motion in a complex, time-varying medium. Such transport processes are denoted as anomalous diffusion and can't be distinguished by the Fick's laws of diffusion. For such problems computer-aided algorithms of Brownian motion often lead to success, implemented as a numerical simulations of discrete random walker processes on networks. We are especially interested in anomalous diffusion on dynamic, complex networks, which alter over time. Propagation on medium can consist of diffusion or of iterative centrality measure such as the Katz prestige, eigenvector centrality or PageRank which all base on random walker processes with distinct behaviors. We try to illustrate the equivalence of (1) the simulation of random walkers, (2) the numerical calculation of the iterative equation and (3) the closed, analytical form of the diffusion processes. As an interdisciplinary background we use networks of clinical history of patients. The nodes of the network stand for the different phenotypes of diseases and the connecting link represents the likelihood that diagnosis appear together. These networks alter in age which is represented in a time-varying, dynamical network. Another main concern is to clarify how far trajectories and probability of presence of random walkers on network correlate with prevalences, evidence and precursor of diseases. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-10-28 14:15 - 15:45 | B. Liu (Medical University of Vienna) | Statistical patterns in decomposed texts
Abstract Different from traditional quantitative linguistics, we study languages in the following ways: distribution of the appearing frequencies in the text and emerging speed of new words. Usually these properties are studied for a book or a specific language, while seldom looks into the other aspects. Thanks to the Corpus of Historical American English(COHA), we get the full text corpus for about more than 110,000 texts ranging from 1810 to 2009. The texts are analyzed using the above approaches in terms of different sentence length. We observe patterns that stay invariant through history. This may indicate the invariance of the habits of language use. But more important, the invariant pattern indicates that the sentence length of 7 is sitting at a special point from the view of new word introduction rate and information density. This result is somehow linked to the results from the famous paper “The Magical Number Seven, Plus or Minus Two”.
| MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-10-21 14:15 - 15:45 | B. Corominas-Murtra (Medical University of Vienna) | Sample Space Reducing Processes: Theory and extensions
Abstract The idea of “path” is fundamental to disentangle the complexity of the systems that are the outcome of an evolutionary process. The path described by an evolving system is often constraining the future space of alternatives, a property usually called history or path-dependence. In this talk I will show how history dependence is deeply linked with the scaling patterns observed in many complex systems, thanks to the recently developed theory of Sample Space Reducing (SSR) processes. From complex networks to critical phenomena, scaling laws emerge in somewhat regular way, and the comprehension of the mechanisms behind scaling patterns has become on of the hot topics of modern statistical physics. SSR process are a totally new route to scaling based on the unique assumption that the phase space reduces as long as the process unfolds. Applications of the theory of SSR processes to diffusive phenomena in complex networks, for example, reveal a source of unexpected results. I will finally revise some work-in-progress extensions of the theory which include i) Sample Space Reducing cascades: Multiplicative processes over a shrinking space, with applications from avalanche study to the energy spectrum of cosmic rays. ii) Sample Space Expanding (SSE) processes, a class of processes which can be understood to be the “mirror” of SSR processes. It is worth to mention that classic results of Record Statistics can be mapped into the simplest case of SSE processes.
| MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-10-14 14:15 - 15:45 | P. Klimek (Medical University of Vienna) | Disentangling genetic and environmental risk factors for diseases from multiplex comorbidity networks Abstract Most disorders are caused by a combination of multiple genetic and environmental factors. If two diseases are caused by the same mechanism, they often co-occur in patients. Here we disentangle how much genetic or environmental risk factors contribute to the pathogenesis of 358 individual diseases, respectively. We pool data on genetic, pathway-based, and toxicogenomic disease-causing mechanisms with co-occurrences obtained from almost two million patients. From this data we construct a multilayer network where nodes represent disorders that are connected by links that either represent phenotypic comorbidity or the joint involvement of certain mechanisms. We quantify the similarity of phenotypic and mechanism-based links for each disorder. Most diseases are dominated by genetic risk factors, while environmental influences prevail for disorders such as depressions, cancers, or dermatitis. The relevance of environmental risk factors for a given disease is inversely related to its broad-sense heritability and also inversely related to the rate at which new drugs for the disease are approved. This might be indicative of a lack of successful drug development for diseases with high environmental risks. Our approach allows to rule out certain types of disease-causing mechanisms when their implied comorbidities are not observed and might therefore be used to identify promising leverage points for the development of future therapies of multifactorial diseases. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-06-24 14:15 - 15:45 | N. Rekabsaz (Vienna University of Technology) | Cancelled | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-06-17 14:15 - 15:45 | A. Hinteregger (University of Vienna) | Network Centrality and Systemic Risk Abstract There are several different measures for the importance of the vertices in a network, often referred to as „Centrality“. I will compare different centrality measures, in particular the Debt-Rank, introduced by Battiston et al. to analyze the systemic risk in a financial network that can lead to a collapse as seen in the financial crises of 2007 - 2008. In our research we try to reconstruct the liabilities between Austrian companies and banks from aggregated data to estimate what companies could lead to a collapse of the economy in the case of a default. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-06-10 14:15 - 15:45 | B. Wang (Medical University of Vienna) | Quantifying systemic risk in a financial and real economy multi-layer network Abstract The financial crises of 2007-2008 has led to increased awareness of the importance of systemic risk, which started pertinent contributions to quantifying the impact of inter-connectedness within financial networks and their interconnection to the real economy. Present research on financial networks and bank-firm networks attempts to estimate systemic risk that is generated by credits or other types of financial contracts. In this work we incorporate ownership-ties as an additional type of interconnection between entities. We generate an Austrian financial and real economy multi-layer network with two layers consisting of liabilities and shares held between and within banks and the real economy. We then quantify the contributions to systemic risk from each layer of this multi-layer network. To generate the network structure we use shareholder data from the Austrian commercial register and aggregated exposure data from banks to the real economy in Austria from the years 2006 to 2015. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-06-03 14:15 - 15:45 | I. Smirnov (Institute of Education, Higher School of Economics, Moscow) | Changing friends is easier than changing grades: evolution of friendship networks of students and social selection Abstract Homophily, the tendency of individuals to associate with others who share similar traits, has been identified as a major driving force in the formation and evolution of social ties. In many cases it is however not clear if an observed homophily is the result of a socialization process, where individuals change their traits according to the dominance of that trait in their local social networks, or if it results from a selection process, in which individuals re-shape their social networks so that their traits match those in the new environment. Here we demonstrate the existence of strong homophily in academic achievements of high school and university students. We analyze a unique longitudinal dataset that contains information about the detailed evolution of a friendship network of 4.500 students across 42 months. Combining the evolving social network data with the time series of the academic performance (gpa) of the individual students, we show that academic homophily is a result of selection: students gradually reorganize their social networks according to their performance levels, rather than adapt their performance to the level of their local group. We are able to understand the underlying dynamics of grades and networks with a simple agent-based model. The lack of a social pull effect in classical educational settings could have important implications for the understanding of the observed persistence of segregation, inequality and limited social mobility in societies. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-05-13 14:15 - 15:45 | H. Fellermann (School of Computing, Newcastle University) | Harvesting biology for computing and production Abstract Molecular biology and microbiology have rendered a picture of living organisms as minutely orchestrated versatile machines and factories that are able to steer flow of matter with unprecedented control and accuracy. In my talk, I will present a portfolio of work that exploits the physical properties of biomolecules and super-molecular biological aggregates for non-biological, technological applications in the areas of molecular computing and material production. In the first part of the talk, I will present recent work on a DNA implementation of a signal recorder based on a stack data structure that is studied in vitro and in simulation. The second part of the talk expands on these principles and uses DNA computing to arrange and control the dynamics of vesicles that are able to carry molecular cargo. By combined molecular and external control mechanisms, we develop a framework to orchestrate bio-chemical synthesis pathways in a manner that mimics the organization of cellular compart-ments such as the Golgi apparatus. Time allowing, I will present some works that integrates the above approaches of refunctionalizing biomolecules to create life-like yet truly non-biological chemical aggregates de novo. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-04-29 14:15 - 15:45 | C. Siebenbrunner (Austrian National Bank, Vienna University of Technology) | Financial Networks and Systemic Risk Abstract My work investigates the implications of systemic risk for financial regulation. I show that in the absence of liquidation costs for defaulted firms, there exists a social dilemma in which socially and pareto efficient bailouts are not performed in a Nash equilibrium by rational agents without government intervention. Next, I establish a general framework for measuring systemic risk capable of separating the effects of direct contagion, asset fire sales and mark-to-market effects. Using Austrian interbank data, I show that the importance of fire sale effects eclipses the effects of the other two channels. Banks-specific indicators of systemic importance are then constructed to assess the efficiency of Basel-III systemic risk regulations, which rely on a simple combination of bank-specific indicators and do not account for network effects. The empirical results suggest that the Basel-III indicator set is an efficient choice of bank-specific indicators, but also that such indicator sets generally perform poorly at capturing the importance of network effects for contagion. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-04-22 14:15 - 15:45 | P. Klimek (Medical University of Vienna) | postponed | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-04-15 14:15 - 15:45 | F. Endel (Vienna University of Technology) | Secondary use of administrative claims data in the Austrian healthcare system: the GAP-DRG database Abstract Secondary use of linked administrative claims data in the healthcare and social insurance system unveils various new possibilities for research and decision making. The Main Association of Austrian Social Security Institutions (“Hauptverband der österreichischen Sozialversicherungsträger”, HVB) operates the so called “GAP-DRG” database, including reimbursement and pseudonymized personal data from the Austrian healthcare system as part of the K-Project DEXHELPP* at the TU Vienna. Two main topics concerning this data collection are addressed. First, current developments as well as the renewal of key components of the database infrastructure in the near future are discussed. Comprehensive changes of hardware and software in the last year – a completely new database server has been deployed – and in the following month are summarized. A novel application platform based on software containers, allowing centralized services and new ways to analyze and present data will be introduced besides improvements due to upgrades of the underlying hardware and software. Second, a concrete example of applied data analysis is presented, bringing together various segments of the database and details of the Austrian public health insurance system. Due to missing genealogical information in GAP-DRG, family relations have to be derived from co-insurance of dependents. Co-insured children and spouses can therefore be distinguished based on their age in relation to the insured person’s age. The presented algorithm can be fitted to diverse objectives concerning the age of parents, spouses and children, while quality control and visualization of multivariate coherences are emphasized.
*http://dexhelpp.at | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-04-08 14:15 - 15:45 | J. Kertesz (Budapest University of Technology and Economics) | Kinetics of Social Contagion Abstract Diffusion of information, behavioral patterns or innovations follows diverse pathways depending on a number of conditions, including the structure of the underlying social network, the sensitivity to peer pressure and the influence of media. We introduce a general model that incorporates threshold mechanism capturing sensitivity to peer pressure, the effect of ‘immune’ nodes who never adopt, and a perpetual flow of external information and study it by analytical methods and simulations. While any constant, non-zero rate of dynamically-introduced spontaneous adopters leads to global spreading, the kinetics by which the asymptotic state is approached shows rich behavior. In particular we find that, as a function of the immune node density, there is a transition from fast to slow spreading governed by entirely different mechanisms. This transition happens below the percolation threshold of network fragmentation, and has its origin in the competition between cascading behavior induced by adopters and blocking due to immune nodes. This change is accompanied by a percolation transition of the induced clusters. We calibrate and validate the model using Big Data. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-03-18 14:15 - 15:45 | R. Hanel (Medical University of Vienna) | Entropy, matrix-multinomials, regular grammars, and processes on multi-graphs - a unifying perspective Abstract Many systems of processes can be characterized by sequences of symbol strings where symbols are sampled from an alphabet (or by sentences of words sampled from a lexicon). Moreover, such symbol sequences may be subject to internal constraints which determine which sequences a process can generate and which not. This can be interpreted as a „grammar'' that distinguishes well-formed sequences from ill-formed ones. For simple grammars (typically referred to as regular grammars) symbols in the underlying alphabet can be represented by matrices acting on a finite dimensional state-space. As a consequence it becomes possible to determine the multiplicity of sequences with identical histograms, and hence the maximum configuration entropy, i.e. the Boltzmann entropy, of the processes. We discuss the general situation in the particular example of the Oslo sandpile-model. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-02-19 14:15 - 15:45 | P. Jizba (Czech Technical University in Prague) | On Statistical Origin of Special and Doubly Special Relativity Abstract Petr Jizba and Fabio Scardigli
In this talk I will show how a Brownian motion on a short scale can originate a relativistic motion on scales that larger than particle's Compton wavelength. I start by discussing complex dynamical systems whose statistical behavior can be explained in terms of a superposition of simpler underlying dynamics — the so-called superstatistics paradigm. Then I go on by showing that the combination of two cornerstones of contemporary physics — namely Einstein’s special relativity and quantum-mechanical dynamics is mathematically identical (when analytically continued to Euclidean regime) to a complex dynamical system described by two interlocked processes operating at different energy scales. The combined dynamic obeys special and doubly special relativity even though neither of the two underlying dynamics does. To model the double-stochastic process in question, I consider quantum mechanical dynamics in a background space consisting of a number of small crystal-like domains varying in size and composition, known as polycrystalline space (or Voronoi tessellation). There, particles exhibit a Brownian motion. The observed relativistic dynamics then comes solely from a particular grain distribution in the polycrystalline space. In the cosmological context such distribution might form during the early universe’s formation. Salient issues such as Hausdorff dimensions of path-integral trajectories, connection with Feynman chessboard model and implications for quantum field theory and cosmology (leptogenesis) will be also briefly discussed.
Related articles:
[1] P. Jizba and F. Scardigli, Eur. Phys. J. C (2013) 73: 2491
[2] P. Jizba and F. Scardigli, Phys. Rev. D (2012) 86: 025029 | MUW, room 513, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2016-02-05 14:15 - 15:45 | I. Smirnov (Institute of Education, Higher School of Economics, Moscow) | Reproduction of Inequality in the Digital Age Abstract The internet and social network sites transformed the world. Instant access to the most of human knowledge and to the millions of peers registered in social networks means that people are not limited by their immediate environment anymore. Can this opportunity be used to break the cycle of inequality reproduction or will we observe the Matthew effect: the rich will get richer and the poor will get poorer? To shed light on this question I've collected a data set from the VK ("Russian Facebook") containing information about 1 million users, their social ties and interests along with their educational background. I'm going to use complex networks analysis in order to estimate the effect of segregation in virtual space and the role of peer influence in overcoming social background. | MUW, Spitalgasse 23, 1090 Vienna, BT86, 3rd floor, seminar room |

2016-01-22 14:15 - 15:45 | B. Fuchs (Medical University of Vienna) | Towards Understanding Collective Social and Economic Dynamics in a Virtual World Abstract Quantitative social science, which strives for a predictive understanding of social facts, has since its origins faced the challenge of limited availability of empirical data. We approach this challenge by turning to a a model-society inside a large-scale virtual world, i.e. an online game, of which we have nearly complete information.
Studying the structure of society as a whole, multiple levels of fractally nested groups and subgroups have been found in real-world anthropological data. By quantitatively analyzing detailed grouping data we find a similar structure in the model-society. This finding indicates that the pattern of fractal organization is universal.
Humans within such a hierarchically organized society are clearly unequal. We contribute to the study of inequality by analyzing the wealth distribution inside the virtual world and by comparing it to real-world observations. Our data allow us to investigate connections between wealth and the position of individuals within their social networks.
We study pairs of interacting individuals and quantify the interplay between physical distance and social interactions. Our findings are compared to previous real-world data. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-12-18 14:15 - 15:45 | N. Rekabsaz (Vienna University of Technology) | The meaning of ‚Meaning‘: Review of the State-of-the-Art in Statistical Word Representation Abstract The endeavor to grasp the meaning of text in Artificial Intelligence starts with understanding the natural building block of the language : „words“. Since years, many methods has provided various computer-understandable representations by trying to resemble the human perception of the meaning and relatedness of the words. The methods basically exploit either human-annotated knowledge or statistical approximation and are broadly used in different fields such as Natural Language Processing, Information Retrieval, Text Mining, Sentiment Analysis, Statistical Machine Translation, etc..
As a main direction for representing words, the statistical methods exploit the implicit knowledge within long text corpora to approximate the representation of words in vector spaces (Word Embeddings). These methods mainly rely on the tendency of natural language to use semantically related words in similar contexts. In the text retrieval community, word and text representation started with Latent Semantic Analysis/Indexing (LSA/LSI), the pioneer approach that initiated a new trend in surface text analysis. Explicit Semantic Analysis (ESA) is one of the early alternatives, aimed at reducing the computational load. However, unlike LSA, ESA does rely on a pre-existing set of concepts, which may not always be available. Random Indexing (RI) is another alternative to LSA/LSI that creates context vectors based on the occurrence of words contexts. It has the benefit of being incremental and operating with significantly less resources while producing similar inductive results as LSA/LSI and not relying on any pre-existing knowledge. Word2Vec further expands this approach using neural networks-based language modeling. When trained on large datasets, it is also possible to capture many linguistic subtleties (e.g., similar relation between Italy and Rome in comparison to France and Paris) that allow basic arithmetic operations within the model.
In this talk, we first review the main word representation methods with focus on Word2Vec as an effective and state-of-the-art approach. Putting forward the question of the meaningfulness of the statistical word representations, we then shed light on the topology of the distributional space in different dimensions, based on recent studies in the IFS group of TU WIEN. The analysis provides an understanding about the informativeness versus randomness of the words' relations and motivates the effective use of the word embeddings in different areas. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-12-04 14:15 - 15:45 | T. Biró (MTA Wigner Research Center for Physics, Hungarian Academy of Sciences) | Entropy: Generalizations, Axioms, Scaling. Abstract We review a particular approach to generalized entropy formulas based on finite reservoir physics. Ideal systems - correlated only due to restrictions of the total phase space - reveal certain leading order corrections to the traditional entropy formula first suggested by Boltzmann and later extended in use and interpretation by Gibbs, Planck and Shannon. We identify which axioms behind the classical formula are violated by this finiteness and how the large dimension scaling may lead to further formulas beyond the well-studied logarithm.
| MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-11-27 14:15 - 15:45 | B. Liu (Section for Science of Complex Systems) | Statistical Properties of English Language Abstract Different from traditional quantitative linguistics, we study languages in the following ways: distribution of the appearing frequencies in the text, emerging speed of new words, nested structure of the potential followers and information needed for creating word sequences among other approaches. Each of the above analyses tells us some properties, either from word-word level or the text as a whole. Usually these properties are studied for a book or a specific language, while other detailed aspects remain vacuum. Thanks to the Corpus of Historical American English(COHA), we get the about more than 110,000 texts ranging from 1810 to 2009 with four genres of fictions, magazines, newspaper articles and non-fictions. The texts are analyzed using the above approaches across history. From the results we observe clear trends in these properties, which indicates that the English language is evolving. Also, texts are decomposed in sentences and then sentences of the same length are clustered and then analyzed. We observe patterns that stay invariant through history. This may indicate the invariance of the habits of language use. The part-of-speech analysis for different sentence length are also presented for the possible explanation of this invariance. These findings may lead to a more comprehensive understand of the statistical properties of language. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-11-13 14:15 - 15:45 | B. Corominas-Murtra (Section for Science of Complex Systems) | Sample Space Reducing processes Abstract The comprehension of the mechanisms behind scaling patterns has become on of the hot topics of modern statistical physics [1]. From complex networks to critical phenomena, scaling laws emerge in somewhat regular way. In this talk I will link the scaling patterns observed in many complex systems with a crucial property behind them: history-dependence [2]. Classical examples of history dependent processes with extremely interesting properties are Pólya urns, the Chinese restaurant or the recurrent random sequences proposed by Ulam and Kac. The link between history-dependence and scaling comes
from the recently defined Sample Space Reducing (SSR) processes [3,4]. SSR processes connect history dependence and scaling in an extremely intuitive way. In addition, SSR process are a totally new route to scaling which can explain a huge range of power-law exponents thanks to the unique assumption that the sampling space is reduced as long as the process unfolds. Simple forms of SSR processes are regular sampling processes where the ‘left-right’ symmetry is broken, leading to a minimal form of history-dependence. In spite of the simplicity of this basic assumption, SSR processes display a wide spectrum of surprising properties. From this zoo of interesting properties, maybe the most remarkable one is the role of the scaling law known as ‘Zipf’s law’ as an attractor, which provides a new, fundamental explanation for the ubiquity of such scaling pattern in real systems. In addition, it provides a new look to diffusive phenomena, since it assumes the presence of a 'target' or 'sink' region, which may model many diffusion-like phenomena, such as urban movement, animal migration or information routing in the internet.
The intuitive rationale behind the SSR processes and the surprisingly simple mathematical apparatus needed to understand them makes the SSR process approach a new research area with promising applications.
Bibliography:
[1] Newman, M E J (2005) “Power laws, Pareto distributions and Zipf's law" Contemporary Physics 46 (5): 323–351.
[2] Arthur, B (1994) Increasing Returns and Path Dependence in the Economy. The University of Michigan Press, Ann Arbor.
[3] Corominas-Murtra, B, Hanel, R, and Thurner, S (2015) “Understanding scaling through history-dependent processes with collapsing sample space”. Proc Nac Acad Sci USA. 112 (17) 5348–5353.
[4] Thurner, S, Hanel, R, Liu, B, and Corominas-Murtra, B (2015) “Understanding Zipf’s law of word frequencies through sample-space collapse in sentence formation”. J R Soc Interface. 12 (108). | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-11-06 14:15 - 15:45 | P. Klimek (Section for Science of Complex Systems) | The forensics of election fraud: statistical detection of voter coercion? Abstract The fundament of a health democracy are free and fair elections. Election forensics is an emerging, interdisciplinary field that aims to develop statistical and data-driven methods to verify whether the preferences of the electorate have been correctly translated into the election outcome. One of the most often reported types of election fraud is voter coercion. This is the practice of intimidating, threatening, or coercing any person for voting or for attempting to vote. Here, we develop a novel statistical method to identify polling centers that show substantial deviations in their electoral behavior with respect to other, geographically close polling centers. We analyse a dataset of twenty different elections in ten different countries and show that in the vast majority of cases no such systematic deviations exist, with the exceptions of electoral referendums after 2003 in Russia, after 2006 in Venezuela, and, to a smaller extent, after 2012 in Mexico. In each of these cases we find a substantial number of polling centers with significantly increased results of turnout and votes for the winning party. Polling centers with a comparably small electorate size are particularly susceptible to such deviations that are compatible with the assumption of widespread voter coercion. We discuss these findings in the context of recent elections in Venezuela, where the deviations are strongest in newly created and strategically located polling centers. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-10-30 14:15 - 15:45 | M. Sadilek (Section for Science of Complex Systems) | Modeling of two-layer dynamics in social systems Abstract Human societies can be viewed as complex systems comprising individuals that are simultaneously active in various social networks, each of the latter representing a different type of social relation.
Examples for such relations are friendship, enmity, communication, or trade. Dynamical changes of a social system's overall structure due to interactions among individuals are accordingly reflected in structural changes of the corresponding social networks.
In the past, some of the observed structural properties of single social networks (e.g. friendship network) could be successfully explained by mathematical models, in the sense that those models generate artificial networks with similar properties. But in general, the individual networks of a social system are not independent from each other: According to structural balance theory, certain triangular configurations of positively connoted relations (such as friendship) and negatively connoted relations (such as enmity) are more balanced that others, and thus more likely to occur.
Using a multilayer network formalism, we are currently developing a mathematical model for the dynamics of social systems with one friendship and one enmity network layer. The model is based on local dynamical rules that implement the key processes preferential attachment and triadic closure, and especially respect structural balance theory.
| MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-10-23 14:15 - 15:45 | R. Hanel (Section for Science of Complex Systems) | Maximum Configuration - Constructing the
Maximum Entropy Principle for Self Reinforcing Processes
Abstract Self reinforcing processes are ubiquitously at work in complex dynamical systems in terms of feed-back loops. Polya urns provide a general framework to study the statistics of such processes and their distribution functions. Maximizing entropy under particular conditions (the maximum entropy principle) allows to predict how likely particular types of systems (weakly interacting components, equilibrium conditions, Markovian, etc.) can be found in particular states. The classical MaxEnt approach however breaks down for non-equilibrium processes such as Polya urn processes. However, using standard maximum configuration considerations we can derive the adequate MaxEnt principle for multi-state Polya urn processes. The form of entropy and constraints to consider in MaxEnt are direct consequences of the class of processes MaxEnt is supposed to apply. This observation has far reaching consequences for an information theory of self reinforcing processes. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-06-26 14:15 - 15:45 | F. Tria (Institute for Scientific Interchange, ISI Foundation Turin) | Dynamics on expanding spaces: modeling the emergence of novelties Abstract Innovation is the driving force in the evolution of human society as well as of biological systems. A general concept that applies to innovation and the emergence of novelties, is what Kauffman called expanding the adjacent possible. The idea is that by creating fresh opportunities, one novelty can pave the way for others, enlarging the space of possibilities in a self-consistent way. I will highlight some statistical features common to many systems where some kind of innovation occurs, and I will shortly review seminal works devoted to their comprehension.
I will then present a recent work aimed at grounding the notion of adjacent possible on real data, by the definition of quantitative measures and the development of a suitable mathematical framework. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-06-19 14:15 - 15:45 | M. Grün (Section for Science of Complex Systems) | A sample-space-reducing approach to study scaling of disease frequencies Abstract Power laws appear in a wide variety of physical, biological and man-made phenomena. They can also be observed in frequencies of human diseases. Over a wide range of magnitudes, rank ordered disease frequencies follow an approximate power-law. Several mechanisms have been proposed to understand the origin of power-law behavior of complex systems. A recently proposed approach offers an alternative way to understand scaling, based on sample space reducing (SSR) processes. In respect to the topology of directed networks, SSR can be related to diffusion processes. It was shown that progression of human diseases can be explained by diffusion on phenotypic disease networks (PDN). I will talk about our studies of diffusion-processes on PDN's and their relation to SSR. I will also discuss if we are able to understand the emergence of scaling in disease frequencies by using a SSR approach. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-06-12 14:15 - 15:45 | C. Flamm (Institute for Theoretical Chemistry, Vienna University) | Computational design of catalysed reaction networks Abstract With the advent of Synthetic Biology, methods for the
computational design of catalysed reaction networks became important. I
will present a computational framework, which combines a graph-grammar base
formalism for describing chemical transformations with optimal flows on
hypergraphs to find solutions for the design problem. An optimal network is
perceived from a set of initial molecules and the set of allowed enzymatic
reactions. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-05-29 14:15 - 15:45 | S. Widder (Department for Computational Systems Biology, Vienna University) | On the structure-function relationship in microbial communities Abstract Classical microbiology has focused on single species. Yet in natural environments and in many industrial applications microbes exist in functional consortia or communities (MCs). MCs play crucial roles in processes as divers as global climate regulation, human health or industrial applications such as wastewater treatment. While sequence-based studies in the last decade revealed immense biodiversity and complexity, we are far from understanding the structure and organisation of MCs and how these translate to function and productivity of microorganisms. Co-occurrence networks of microorganisms from environmental samples facilitate understanding and analysis of this structure-function relationship. In my talk I will focus exemplarily on one ecological and one pathological MCs and their organisation. I will show how the analysis of MC network topology allows inference of functional roles in riverine biofilm MCs and demonstrate that environmental perturbation leads to network fragmentation. For the pathologic MC (cystic fibrosis lung microbiota) we could confirm by network analysis that a shift in the severeness of the condition is related to a taxonomic shift in the MC and the related metabolic core processes. Furthermore we propose functional and taxonomical keystones as targets for novel drug development. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-05-22 14:15 - 15:45 | M. Sadilek (Medical University of Vienna) | Models for social networks with negative ties Abstract Social balance theory suggests that within a social network, the subnetworks comprising positive ties (friendship network) respectively negative ties (enmity network) are dynamically interrelated. I will present some results from my recent work which concentrates on finding a realistic generative model for such networks based on data from the massive multiplayer online game Pardus. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-04-29 14:00 - 15:00 | G.C. Rodi (Institute for Scientific Interchange - ISI Foundation Turin) | Optimal learning paths in information networks Abstract Each sphere of knowledge and information could be depicted as a complex mesh of correlated items. By properly exploiting these connections, innovative and more efficient navigation strategies could be defined, possibly leading to a faster learning process and an enduring retention of information. It has been investigated how the topological structure embedding the items to be learned can affect the efficiency of the learning dynamics. To this end we introduce a general class of algorithms that simulate the exploration of knowledge/information networks standing on well-established findings on educational scheduling, namely the spacing and lag effects. While constructing their learning schedules, individuals move along connections, periodically revisiting some concepts, and sometimes jumping on very distant ones. In order to investigate the effect of networked information structures on the proposed learning dynamics we focused both on synthetic and real-world graphs such as subsections of Wikipedia and word-association graphs. We highlight the existence of optimal topological structures for the simulated learning dynamics whose efficiency is affected by the balance between hubs and the least connected items. Interestingly, the real-world graphs we considered lead naturally to almost optimal learning performances. | Section for Science of Complex Systems |

2015-04-24 14:15 - 15:45 | G. Tkacik (Institute of Science and Technology Austria) | Critical behavior in networks of real neurons Abstract The patterns of joint activity in a population of retinal ganglion cells encode the complete information about the visual world, and thus place limits on what could be learned about the environment by the brain. We analyze the recorded simultaneous activity of more than a hundred such neurons from an interacting population responding to naturalistic stimuli, at the single spike level, by constructing accurate maximum entropy models for the distribution of network activity states. This – essentially an ”inverse spin glass” – construction reveals strong frustration in the pairwise couplings between the neurons that results in a rugged energy landscape with many local extrema; strong collective interactions in subgroups of neurons despite weak individual pairwise correlations; and a joint distribution of activity that has an extremely wide dynamic range characterized by a zipf-like power law, strong deviations from ”typicality”, and a number of signatures of critical behavior. We hypothesize that this tuning to a critical operating point might be a dynamic property of the system and suggest experiments to test this hypothesis. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-04-16 14:00 - 15:00 | J.-P. Onnela (Harvard University, Boston) | Cell phones, Social Networks, Crowds, and Digital Phenotyping Abstract Cell phones are now ubiquitous: it is estimated that the number of phones in use exceeds the size of the global population in 2015. I will talk about two main lines of work in our lab. First, we use call detail records (CDRs) to investigate the structure of large-scale social networks and their relationship to underlying geography. We also use these data to study population dynamics at massive gatherings of people and to inform models of epidemic spread, in particular cluster randomized trials. Second, we have coined the term digital phenotyping to refer to the moment-by-moment quantification of the individual-level human phenotype, in situ, using data from digital devices. Using a dedicated smartphone application, we collect active data (surveys, voice samples, etc.) and passive data (spatial location, social engagement, etc.) from consenting patients and analyze these data using modern statistical learning methods. I will talk about some of our work that utilizes these two approaches, what types of insights they may yield to the study of social networks and human behavior, and how this work interfaces with public health. | MUW, Jugendstil seminar room, Spitalgasse 23, 1090 Vienna, BT88, 2nd floor |

2015-04-10 14:15 - 15:45 | J. Menche (Center for Complex Networks Research and Department of Physics, Northeastern University) | Diseases in the human interactome Abstract Recent advances in disease gene identification and high-throughput mapping of physical interactions between gene products offer new opportunities to explore the role of molecular networks in human disease. Here we show that proteins associated with the same disease display a statistically significant tendency to agglomerate in the same neighborhood of the interactome, offering quantitative evidence for the existence of well-localized and potentially identifiable disease modules. Most important, we find that the network-based location of each disease module determines its pathobiological relationship to other diseases. For example, disease pairs with overlapping modules show significant co-expression patterns, symptom similarity, and comorbidity; those that reside in separated network neighborhoods are pathobiologically and clinically distinct. The proposed interactome-based framework offers systematic avenues to discover common molecular roots between clinically unrelated disease phenotypes even if they do not share disease genes, and helps identify the biological role of GWAS genes of small effect size and low genome-wide significance. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-03-20 14:15 - 15:45 | I. Kondor (Parmenides Foundation, Pullach b. München and Corvinus University Budapest ) | Risk measures, regularization and market impact Abstract This talk will briefly review the history of risk measures appearing in the subsequent generations of international banking regulation, their relative merits and shortcomings, with special emphasis on Expected Shortfall (ES) that is on its way to becoming the next global regulatory market risk measure. When these measures are used to predict the out-of-sample risk or to optimize portfolios they all display a weakness due to the relative scarcity of data compared to the size of institutional portfolios. The resulting estimation errors can be very large, in fact, for some critical values of the parameters they can diverge, with the optimization algorithm undergoing a phase transition. The estimation error problem is particularly serious for downside risk measures, such as ES. The standard way to tackle these high dimensional problems is to use regularization, that is to introduce a penalty for excessive positions. It is shown that investors who optimize their portfolios under any of the coherent risk measures are naturally led to regularized portfolio optimization when they take into account the impact their trades make on the market. The impact function determines which regularizer is to be used. It is shown that any regularizer based on the norm Lp with p > 1 makes the sensitivity of coherent risk measures to estimation error disappear, while regularizers with p < 1 do not. The L1 norm represents a border case: its “soft” implementation does not remove the instability, but rather shifts its locus, whereas its “hard” implementation (equivalent to a ban on short selling) eliminates it. We demonstrate these effects on the important special case of Expected Shortfall. | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-03-06 14:15 - 15:45 | R. Hanel (Section for Science of Complex Systems) | How to gauge prior probabilities for hypotheses testing and obtain
critical threshold landscapes of the correlation coefficient
Abstract Lindley's paradoxon is an effect that haunts Bayesian Hypotheses testing
especially if hypotheses such as H1: A and B are correlated, versus H0: A
and B are independent, are tested against each other. There is of course no real paradoxon only prior probabilities that have been chosen badly. We show that by a simple technique, comparing extreme samples which represent the hypotheses in their purest obtainable form, allows to gage the prior probabilities of the hypotheses adequately.
We analyze the situation in the context of binary processes and show that in
the limit of large numbers of samples the posterior probabilities of the
hypotheses H1 and H0 exhibit a first order phase transition at critical
values of the correlation coefficient.
| MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2015-02-11 10:00-12:00 | J.S. Lansing (Complexity Institute at Nanyang Technological University in Singapore) | Genetic and Language Trees in Indonesia | Section for Science of Complex Systems |

2015-01-23 14:15 - 15:45 | M. Sadilek (Section for Science of Complex Systems, MUW) | Co-evolution of networks | MUW, Spitalgasse 23, 1090 Vienna, BT86, seminar room, 3rd floor |

2015-01-16 14:15 - 15:45 | G. Percino-Figueroa (Section for Science of Complex Systems, MUW) | Linking dynamics of tensored networks in virtual society | MUW, Spitalgasse 23, 1090 Vienna, BT86, seminar room, 3rd floor |

2015-01-09 14:15 - 15:45 | S. Aichberger (Section for Science of Complex Systems) | Medical multiplex networks | MUW, Spitalgasse 23, 1090 Vienna, BT86, seminar room, 3rd floor |

2014-12-05 14:15 - 15:45 | B. Corominas-Murtra (Section for Science of Complex Systems, MUW) | Sample space reducing processes | MUW, Informatics library, Spitalgasse 23, 1090 Vienna, BT88, 3rd floor |

2014-11-28 14:15 - 15:45 | R. Hanel (Section for Science of Complex Systems, MUW) | Molecular regulatory networks and constraints from a dynamical systemic perspective | MUW, Spitalgasse 23, 1090 Vienna, BT86, seminar room, 3rd floor |

2014-11-21 14:15 - 15:45 | P. Klimek (Section for Science of Complex Systems, MUW) | Systemic Trade Risk | MUW, Informatics library, Spitalgasse 23, 1090 Vienna,BT88, 3rd floor |

2014-10-31 14:15 - 15:45 | B. Fuchs (Section for Science of Complex Systems, MUW) | Physical potentials derived from human interactions | MUW, Spitalgasse 23, 1090 Vienna, BT86, seminar room, 3rd floor |

2014-10-17 14:15 - 15:45 | I. Daruka (Johannes Kepler University Linz) | A sustainable avalanche? Publication dynamics and the proliferation of research journals Abstract The number of research papers published yearly shows a staggering exponential growth. One can in fact witness a century-long publication avalanche. According to the two major databases INSPEC and Thomson Reuters, there is a 300-fold increase in the number of published items since the year 1900. The world of science has recently reached such a level of proliferating complexity that its structural evolution itself poses challenging scientific problems. | Medizinische Universität, Bauteil 88, Bibliothek 'Zentrum für Medizinische Statistik, Informatik, und Intelligente Systeme', Ebene 3 |

2014-06-27 14:15 - 15:45 | K. Dovzhik (University of Vienna) | Phase transitions in collective intelligence systems | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2014-06-20 | NO SEMINAR | ||

2014-06-13 14:15 - 16:15 | Lechner; Mitterwallner (University of Vienna) | Quantum phase transitions | Universität Wien, Fakultät für Physik, E. Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2014-06-06 | NO SEMINAR | ||

2014-05-30 | NO SEMINAR | ||

2014-05-23 14:15 - 15:45 | M. Formanek (University of Vienna) | Phase transitions in telechelic star polymers Abstract Novel colloidal systems constitute a broad and developing new field of research in soft matter science. Telechelic star polymers (TSPs), which are a hybrid system that combines the flexibility of chains with the sphericity of colloids, are particularly interesting materials because of their high degrees of freedom and the complex structural and dynamical properties that arise from them. This talk will focus on the ability of TSPs to self-assemble, to undergo macroscopic as well as microscopic phase transitions and to form multiply-connected percolating networks, depending on their functionality and the ratio of solvophilic to solvophobic groups. | Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2014-05-16 14:15 - 15:45 | E. Traxler (University of Vienna) | Phase transitions in large networks Abstract The concept of networking is used in many applications: computational processes (e.g., neural networks), transport, ecology, economics, sociology and many others. Therefore, understanding the dynamics of large, complex networks is becoming increasingly important. In these systems, stability is of utmost importance. Instabilities always lead to disaster and can even assume global proportions. Whereas stable systems can regenerate themselves and thus have some degree of fault tolerance, making them partially immune even to attacks. The question of the conditions for stability in networks is a central theme of my lecture. | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2014-05-09 14:15 - 15:45 | M. Sadilek (Section for Science of Complex Systems, CeMSIIS, MUW) | Phase transitions in the Kuramoto model | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2014-04-11 14:15 - 15:45 | D. Lin (University of Vienna) | Phase transition in random catalytic networks Abstract Phase transitions are one of the most interesting topics in physics. They occur in various physics-related disciplines from thermodynamics to network theory. In my talk I'd like to present the content of the paper "Phase transition in random catalytic networks" (by Rudof Hanel, Stuart A. Kauffman and Stefan Thurner). I'll start with some usefull definitions for the description of catalytic networks (like the "product rule density", the "support of a process", the "forward difference", etc.). Afterwards I'll derive the growth equation for catalytic sets and an analytical approximation of the forward-closure size. Plotting the "final number of products" against the "pair density" will lead to diagrams equal to PV-diagrams when considering Van der Waals equation and therefore we'll identify phase transition in catalytic networks. | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2014-04-04 14:15 - 15:45 | D. Grumiller (Faculty of Physics, Vienna University of Technology) | Cooking recipe for a Universe Abstract Phase transitions are ubiquitous in Nature. Interestingly, even spacetime itself can be subject to such transitions, like in the famous Hawking-Page phase transition between 'hot curved space' and certain black holes. Recently, we found a similar phase transition between 'hot flat space' and an expanding cosmological spacetime in 2+1 dimensions. I review the ingredients contained in our 'cooking recipe' for a Universe. | Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2014-03-28 | NO SEMINAR | ||

2014-03-21 14:15 - 15:45 | P. Schuster (Faculty of Chemistry, University of Vienna) | Phase Transitions in Evolution: When do quasispecies form error thresholds? Abstract The notion of phase transitions is frequently used as a metaphor for abrupt changes in evolutionary dynamics. Here it will by understood more rigorously in the context of selection mutation dynamics. Replication and mutation are modeled as parallel chemical reactions, and evolutionary dynamics is seen as an adaptive or a random walk of population in an abstract space of genotypes represented by nucleic acid sequences. Depending on the distribution of fitness values over sequence space, populations can tolerate replication errors only up to a certain maximal mutation rate, which has been characterized as an error threshold. Below threshold and in the long time limit the populations converge to stationary sequence distributions denoted as quasispecies, whereas above threshold the populations are permanently non-stationary and drift randomly through sequence space. Accordingly, error thresholds are transitions from localized quasispecies to migrating populations, and they become sharper and sharper with increasing lengths of sequences. They have much in common with phase transitions. In the seminar we shall explore the conditions for the existence of error thresholds and analyze the nature of the transitions. | Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2014-03-14 14:15 - 15:45 | H. Grosse (Faculty of Physics, University of Vienna) | From models describing phase transitions to QFT Abstract I first describe simple statistical physics models discribing phase transitions in various dimensions. Ferromagnetic ones are related to lattice regularized Euclidean field theory.
In dimensions greater than four one obtains triviality.
Next I describe renormalization group ideas and mention the Landau ghost problem.
An improvement is obtained by deforming space-time. Together with Raimar Wulkenhaar we found a model which is asymptotic safe and which was constructed recently. The connection to matrix models is mentioned. | Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2014-03-07 14:15 - 15:45 | W. Pueschl (Faculty of Physics, University of Vienna) | Basics of phases and phase transformations: An introductory tour from the viewpoint of a materials physicist Abstract In this overview, we start from the notion of a phase and ask ourselves why it is important to understand and distinguish phases, which leads us to the high cultural and technological significance of multi-phase structures. The equilibrium states of phases such as displayed in phase diagrams can be sought by means of thermodynamic potentials. We discuss how to calculate the quantities found therein. Regarding two-component (as the simplest case of multi-component) systems, it is seen how phase diagrams can easily be derived by graphical thermodynamics. Thermodynamics requires that on entering a miscibility gap a new phase be generated. This leads us to precipitation as a particularly interesting case of phase transformation, with nucleation and spinodal decomposition as alternative pathways.
Opposed, in a certain sense, to phase separation are ordering transitions. Atomic configuration can be described by various order parameters based on correlation functions. We consider short-range order and long-range order. The Bragg-Williams model of ordering is the simplest ansatz leading to meaningful qualitative behavior. We recognize in it a one-point mean field approximation to the Ising model. In the framework of the Ising model ordering / phase- separating alloys can be seen analogous to magnetic systems.
Finally, martensitic phase transformations are a typical solid-state phenomenon. Once initiated, they proceed almost instantaneously, practically at the speed of sound, and allow remarkable applications such as shape-memory alloys.
| Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2014-01-31 14:15 - 15:45 | W. Dvorak (Forschungsgruppe Theory and Applications of Algorithms, University of Vienna) | Uniform Price Strategies to Exploit Positive Network Externalities. Joint work with: Ludek Cigler, Monika Henzinger, Martin Starnberger Abstract Assume a seller wants to sell a digital product in a social network where nodes represent clients and edges represent a friend-like relationship between them. Assume further that the value of the item for a buyer has positive network externalities, i.e., it is a monotone function of the set of neighbors that already have the item. The goal of the seller is to maximize his revenue. Previous work on this problem studies the case where clients are offered the item in sequence and have to pay personalized prices. This is highly infeasible in large scale networks such as the Facebook graph: (1) Offering items to the clients one after the other consumes a large amount of time, and (2) price-discrimination of clients could appear unfair to them and result in negative client reaction or could conflict with legal requirements. On the other side, offering the same price to every client significantly reduces the seller’s revenue.
Thus, we study settings where we limit the duration of the selling process and the amount of price discrimination. Specifically, the item is offered in parallel to multiple clients at the same time and at the same price. This is called a round. We show that with O(log n) rounds, where n is the number of clients, a constant factor of the revenue with price discrimination can be achieved and that this is not possible with o(log n) rounds. Moreover we show that it is APX-hard to maximize the revenue even if the externalities in the clients’ valuation functions are very restricted, and we give constant factor approximation algorithms for various further settings of limited price discrimination.
| Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2014-01-24 14:15 - 15:45 | P. Klimek (Section for Science of Complex Systems, CeMSIIS, MUW) | Systems Medicine& Big Data - New Frontiers in Medicine | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2013-12-13 14:15 - 15:45 | M. Sadilek (Section for Science of Complex Systems, CeMSIIS, MUW) | Models of Synchronization in Networks | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2013-12-06 14:15 - 15:45 | A. Chmiel (Section for Science of Complex Systems, CeMSIIS, MUW) | Manlio De Domenico et.al: Mathematical formulation of multi-layer networks Abstract A network representation is useful for describing the structure of a large variety of complexsystems. However, most real and engineered systems have multiple subsystems and layers of connectivity,and the data produced by such systems is very rich. Achieving a deep understandingof such systems necessitates generalizing \traditional" network theory, and the newfound deluge of data now makes it possible to test increasingly general frameworks for the study of networks.In particular, although adjacency matrices are useful to describe traditional single-layer networks,such a representation is insucient for the analysis and description of multiplex and time-dependent networks. One must therefore develop a more general mathematical framework to cope with the challenges posed by multi-layer complex systems. In this paper, we introduce a tensorial framework to study multi-layer networks, and we discuss the generalization of several important network descriptors and dynamical processes|including degree centrality, clustering coecients, eigenvector centrality, modularity, Von Neumann entropy, and diusion|for this framework. | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2013-11-29 14:15 - 15:45 | B. Corominas-Murtra (Section for Science of Complex Systems, CeMSIIS, MUW) | The core structure of complex networks Abstract What is the core of a complex network? In this talk we will discuss
this crucial question of modern complex networks theory. To start
with, we will briefly revise the underlying theory of network
decomposition and nested subgraphs. We will then discuss specific
core-finding and network decomposition methods. Special emphasis will
be given to the K-core, the paradigmatic method of core decomposition.
It is worth to note that the K-core is nowadays a source of both
mathematical and physical problems, but that it was proposed more than
30 years ago to disentangle the deep structure of social networks.
However, this core decomposition mechanism has several limitations
that lead to new, more sophisticated methods, like the GK-core or the
M-core, which better agree to the notion of core in specific problems,
specially those related to social networks. We will finally revise
some applications of this formal/algorithmic machinery to real
problems, like the understanding of the network dynamics defined by
virtual society of the massive on line game 'Pardus'.
[1] Seidman, Stephen B. (1983), "Network structure and minimum
degree", Social Networks 5 (3): 269–287
[2] Dorogovtsev, S.N.; Goltsev, J.F.F.; Mendes, JF (2006), "k-core
organization of complex networks", Physical Review Letters 96 (4):
040601
[3] B Corominas-Murtra, JFF Mendes, RV Solé (2008) "Nested subgraphs
of complex networks",
Journal of Physics A: Mathematical and Theoretical 41 (38), 385003
[4] Bernat Corominas-Murtra, Benedikt Fuchs, Stefan thurner "B
Corominas-Murtra, B Fuchs, S Thurner" arXiv preprint
http://arxiv.org/abs/1309.6740
[5] Pol Colomer-de-Simón, M. Ángeles Serrano, Mariano G. Beiró, J.
Ignacio Alvarez-Hamelin, Marián Boguñá "Deciphering the global
organization of clustering in real complex networks", Scientific
Reports 3, Article number: 2517 doi:10.1038/srep02517 | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2013-11-22 14:15 - 15:45 | J.D. Whitfield (Faculty of Physics, University of Vienna) | Quantum to classical transitions: application to continuous time Markov processes Abstract I will give an detailed introduction to Markov theory in quantum and classical contexts and its relation to the "measurement problem" in quantum mechanics. This is followed by an introduction to some notions in graph theory and random walks. The talk will then centered on our article "Quantum Stochastic Walks" Phys. Rev. A, vol 81 page 022323 (2010) and the axiomatic introduction of the quantum stochastic walk. We will then discuss related results on thermodynamics, directed quantum transport, and Lieb-Robinson bounds. | Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2013-11-15 14:15 - 15:45 | S. Poledna (Section for Science of Complex Systems, CeMSIIS, MUW) | Controlling Financial Networks | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2013-11-08 13:00 - 14:00 | A. Bazzani (National Institute of Physics and Astronomy, University of Bologna) | Statistics of human mobility in Italian traffic data | Medical University of Vienna, Complex Systems Group, seminar room, Spitalgasse 23, 1090 Vienna |

2013-11-08 14:15 - 15:45 | B. Baumgartner (University of Vienna) | Irreversible Quantum Dynamics | Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2013-10-18 | R. Hanel (Section for Science of Complex Systems, CeMSIIS, MUW) | Maximum Entropy for Aging Complex Systems Abstract Statistical mechanics of many statistical systems allows to asymptotically replace expectation values by the so called `maximum configuration'. Sampling processes, where the probabilities of observing a system in particular states evolves stationary, can be described by the reduced Boltzmann entropy that in such cases takes the functional form of Shannon entropy.
The probability of observing a non Markovian, aging system in a certain state is by definition not a stationary process and Boltzmann entropy of such systems becomes non extensive. The `maximum configuration' argument on the other hand still holds also for sufficiently large aging systems.
Using that entropy is a function that only depends on the frequencies of observing the system in a particular state allows to show that the reduced Boltzmann entropy of aging systems takes the functional form of generalized entropies, as for instance Tsallis entropy.
| Universität Wien, Fakultät für Physik, Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 1090 Wien |

2013-10-04 14:15 - 15:45 | K. Hlavackova-Schindler (Bioinformatcis, Boku University Vienna) | Equivalence of Granger Causality and Transfer: A Generalization Abstract Barnett et al. in 2009 proved that Granger causality and transfer entropy causality measure are equivalent for time series which have a Gaussian distribution. Granger causality test is linear, while transfer entropy a non-linear test. Many biological and physical mechanisms show to have non-Gaussian distributions. In this paper we investigate under which conditions on probability density distributions of the data can the equivalence of the two causality measures be extended. In the complexity sense ”cheaper” linear Granger test can be applied for detection of causality in time series satisfying these conditions. These results have an impact on causality detection in common biological and physical time series. | Medical University of Vienna, Informatics Library, Spitalgasse 23, 1090 Vienna, BT88, E03 |

2013-03-05 13:00 - 14:00 | S. Thurner (Section for Science of Complex Systems) | Preliminary meeting to "Complex Systems II - Applications" lecture & exercise | Informatics Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2012-07-27 12:15 - 13:45 | K. Goh (Dept. of Physics, Korea University, Seoul) | Processes on Multiplex Networks | Seminar room COSY, 1090 Vienna, Spitalgasse 23, BT86 E03 |

2012-06-29 14:00-15:30 | G. Kelnhofer (Universität Wien) | Thermal quantum (gauge) field theory | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität WIen, Boltzmanngasse 5/5.St., 1090 Wien |

2012-06-22 14:15 -15:45 | C. Tsiapalis (Section for Science of Comlex Systems, Medical University of Vienna) | Citizen Science | Informatics Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2012-06-15 | H. Hüffel (Fakultät für Physik, Universität Wien) | Quantum (gauge) field theory on bipartite lattices | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität WIen, Boltzmanngasse 5/5.St., 1090 Wien |

2012-06-01 14:15 -15:45 | M. Schoelling (Section for Science of Comlex Systems, Medical University of Vienna) | How deterministic is gene transcription? | Informatics Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2012-05-11 14:15 -15:45 | B. Fuchs (Section for Science of Comlex Systems, Medical University of Vienna) | Basic Properties of Complex Networks | |

2012-05-04 14:00-15:30 | L. Hingerl (Universität Wien) | Unaboidable order in nonequilibrium systems | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität WIen, Boltzmanngasse 5/5.St., 1090 Wien |

2012-04-27 14:15 - 15:45 | P. Klimek (Section for Science of Complex Systems, Medical University of Vienna) | Linear Response Theory | |

2012-04-20 14:15 -15:45 | S. Poledna (Section for Science of Comlex Systems, Medical University of Vienna) | What is Fisher Information? | |

2012-03-30 14:00-15:30 | F. Harbich (Universität Wien) | White noise limits of multiplicative colored noise | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität WIen, Boltzmanngasse 5/5.St., 1090 Wien |

2012-03-23 14:00-15:30 | M. Sadilek (Universität Wien) | Spectral analysis of white noise | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität WIen, Boltzmanngasse 5/5.St., 1090 Wien |

2012-03-16 14:00 - 15:00 | E. Feireisl (Institute of Mathematics, Academy of Sciences of the Czech Republic, Prague; ESI Wien) | Mathematics of complete fluid systems | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität WIen, Boltzmanngasse 5/5.St., 1090 Wien |

2012-03-01 13:00 | Vorbesprechung LVs SS 2012 (Institut für Wissenschaft Komplexe Systeme, Medizinische Universität Wien) | Seminarraum Inst. f. Wiss. Kompl. Systeme, 9., Spitalg. 23, BT 86, 3. St. | |

2012-01-27 14:15 -15:45 | B. Fuchs (Section for Science of Complex Systems, Medical University of Vienna) | Progress Report on Group Formation and Dynamics in the Pardus Game | Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2012-01-20 14:15-15:45 | R. Hanel (Section for Science of Complex Systems, Medical University of Vienna) | Entropy of a Non-Extensive Spin Model | Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2011-12-16 14:00-15:30 | M. Schoelling (Section for Science of Complex Systems, Medical University of Vienna) | Reasoning about cellular diversity within a sequentially linear regulatory network model | Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2011-12-02 14:00-15:30 | M. Szell (Section for Science of Complex Systems, Medical University of Vienna) | Recent Developments in Socio-dynamics | Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2011-11-25 14:00-15:30 | S. Poledna (Section for Science of Complex Systems, Medical University of Vienna) | The Role of Leverage in a World of Perfect Hedging Abstract We use a toy model of the financial market to test the efficiency and dangers of credit regulation schemes. We find that Basle-type regulation works fine in situations of low leverage levels in the financial system, however they become destabilizing in scenarios with realistic leverage level. We further design an ideal world, where all leverage introduced risk is hedged with options. Even by assuming that option writers never default, we see that introducing the heavy requirement of complete hedging does not make the system systemically more secure. | Library, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2011-11-04 14:15 - 15:45 | P. Klimek (Section for Science of Complex Systems, Medical University of Vienna) | Empirical Confirmation of Creative Destruction | Seminar room no. 513, CeMSIIS, Medical University of Vienna, building 88, level 3 |

2011-10-25 14:00-15:30 | Vorbesprechung (Institut für Wissenschaft Komplexe Systeme, Medizinische Universität Wien) | Lehrveranstaltungen WS 2011/12 | Seminarraum Inst. f. Wiss. Kompl. Systeme, 9., Spitalg. 23, BT 86, 3. St. |

2011-07-01 14:30-16:00 | R. Sinatra (Section for Complex Systems, Medical University of Vienna) | Entropy rate of random walks on networks Abstract
In the last decade increasing attention has been devoted to the study of random walks on complex topologies. Various features of random walks on networks, such as passage times and spectral properties have been investigated, and random walks have also been used to to detect communities, to evaluate centrality of nodes and to coarse-grain graphs. Another quantity recently considered for the study of random walks is the entropy rate, a measure used to characterize the mixing properties of a stochastic process.
In this talk we will first discuss some of the properties of random
walks which might have many relevant applications. In particular, we will consider biased random walks, i.e. random walks with a jumping probability which depends on some properties of target node. We will then focus on designing biased random walks with maximal entropy rate on a given graph, i.e. choosing the transition probabilities of the random walk in such a way that the random walkers are maximally dispersing in the graph, exploring every possible walk with equal probability. Although in principle the optimization of the entropy rate requires that a walker has, at each time step, a global information on the structure of the entire graph, information which is often unavailable, we will demonstrate that it is always possible to construct maximal-entropy random walks by defining a set of transition probabilities that are markovian and that rely only local information on the graph structure. | Seminar room no. 513, CeMSIIS, Medical University of Vienna, building 88, level 3, Spitalgasse 23, 1090 Vienna |

2011-06-17 14:30-16:00 | R. Winkler (Institut für Diskrete Mathematik, Technische Universität Wien) | Mittelwerte sind wahrscheinlich, doch Extrema sind typisch - Bemerkungen zum Antagonismus von Maß und Kategorie Abstract Zentrale Teile der Stochastik kreisen um das Gesetz der großen Zahlen, wonach (bereits unter schwachen Voraussetzungen) die Konvergenz von Mittelwerten gegen den Erwartungswert fast sicher ist. Aus topologischer Sicht, d.h. im Sinne Bairescher Kategorien, ist jedoch ein gänzlich konträres Verhalten typisch. Der allgemeine mathematische Hintergrund dieses Phänomens soll ebenso beleuchtet werden wie einige charakteristische Beispiele.
| Medizinische Universität, Bauteil 88, Seminarraum 513, Ebene 3, Spitalgasse 23, 1090 Wien |

2011-06-10 14:30-16:00 | N.N. (Institut für Wissenschaft Komplexe Systeme, Medizinische Universität Wien) | Studenten-Kurzvortrag | Bibliothek 'Zentrum für Medizinische Statistik, Informatik, und Intelligente Systeme', Medizinische Universität Wien, Bauteil 88, Ebene 3 |

2011-06-03 14:30-16:00 | N.N. (Institut für Wissenschaft Komplexe Systeme, Medizinische Universität Wien) | Studenten-Kurzvortrag | Bibliothek 'Zentrum für Medizinische Statistik, Informatik, und Intelligente Systeme', Medizinische Universität Wien, Bauteil 88, Ebene 3 |

2011-05-27 14:30-16:00 | N.N. (Institut für Wissenschaft Komplexe Systeme, Medizinische Universität Wien) | Studenten-Kurzvortrag | Bibliothek 'Zentrum für Medizinische Statistik, Informatik, und Intelligente Systeme', Medizinische Universität Wien, Bauteil 88, Ebene 3 |

2011-05-20 14:30-16:00 | Ch. Likos (Fakultät für Physik, Universität Wien) | Electrostatics and soft matter: from star-branched polyelectrolytes to patchy colloids | Erwin Schrödinger Hörsaal, Fakultät für Physik, 9., Boltzmanngasse 5, 5.Stock |

2011-05-13 14:00-15:30 | B. Kuzmany (Doktoratskolleg Galizien, Universität Wien) | Physik und Technologie von Graphen | Medizinische Universität, Bauteil 88, Bibliothek 'Zentrum für Medizinische Statistik, Informatik, und Intelligente Systeme', Ebene 3 |

2011-05-06 14:30-16:00 | C. Dellago (Fakultät für Physik, Universität Wien) | Studying rare events with transition path sampling | Erwin Schrödinger Hörsaal, Fakultät f. Physik, Universität Wien, 9., Boltzmanngasse 5, 5. Stock |

2011-04-29 14:30-16:00 | H. Posch (Computergestützte Physik, Universität Wien) | Stochastic and dynamical computer thermostats | Medizinische Universität, Bauteil 88, Bibliothek 'Zentrum für Medizinische Statistik, Informatik, und Intelligente Systeme', Ebene 3 |

2011-04-15 14:30-16:00 | P. Walther (ÖAW, Inst. f. Quantenoptik u. Quanteninformation) | Photonic quantum simulation of frustration in chemical and physical systems | Universität Wien, Fakultät für Physik Erwin Schrödinger Hörsaal, Boltzmanngasse 5 / 5. Stock, 1090 Wien |

2011-04-01 14:00-15:30 | K. Temme (Universität Wien) | Thermal States on a Quantum Computer | Medizinische Universität, Bauteil 88, Seminarraum 513, Ebene 3 |

2011-03-18 14:30-16:00 | R. Hanel (Section for Complex Systems) | Generalized Entropy and Extensivity Abstract A short introduction to the concept of generalized entropies: The necessity of generalized entropies is not primarily grounded in alternative means to reproduce particular types of observed distribution functions from some modified maximum entropy principle - but by the concept of Extensivity.
By relaxing the axioms of Information Theory an axiomatic approach towards generalized entropies can be taken. The generic functional form of generalized entropies can be deduced as well as conditions which allow to determine which entropy will be the extensive entropy of some system - given the knowledge of how phase-space expands as a system is enlarged.
| |

2011-03-17 15:00 | M. Schoelling (Section for Complex Systems) | Exact Matrix Completion Using the Singular Value Threshold Algorithm | |

2011-03-04 11:00 | Lehrveranstaltungen SS 2011 Vorbesprechung (Section for Complex Systems) | Seminar room BT 86, E 03, Spitalgasse 23, 1090 Vienna | |

2011-02-04 14:00-15:00 | A. Lörincz (Eötvös Loránd Univ., Dept. of Software Technol. a. Methodol.) | Tools for CHI: From face tracking and reinforcement learning to optimization of typing tool Dasher | Seminarraum BT 86, Ebene 3 |

2011-01-28 14:30-16:00 | S. Poledna (Medical University of Vienna) | Danger of leverage in a world of perfect hedging | Medizinische Universität, Bauteil 88, Seminarraum 513, Ebene 3 |

2011-01-14 14:00-15:30 | M. Schoelling (Medical University of Vienna) | Gene Expression Regulation - from DNA to RNA Abstract The regulation of gene expression in live-forms is subject to high
> dynamics and nonlinearities.
> To create a mathematical model of these systems biological processes
> must be reflected adequately.
> This talk will handle the biological foundations of gene expression from
> DNA to RNA including eukaryotic characteristics like RNA splicing with
> respect to the influence on the mathematical model.
| Medizinische Universität, Bauteil 88, Seminarraum 513, Ebene 3 |

2011-01-07 14:30-16:00 | M. Poechacker (Medical University of Vienna) | Stability by adapted dimensionality of dynamics in a minimally non-linear model of gene regulation | |

2010-12-17 14:30-16:00 | M. Szell (Medical University of Vienna) | Mobility of human-controlled characters on a synthetic network Abstract We study long-time mobility of human-controlled characters on a network-shaped universe of a massive multiplayer online game. We take a number of mobility measurements and compare them with measures of simulated random walkers on the same topology. Mobility of players is sub-diffusive - the mean squared displacement follows a power law with exponent 0.4 - and significantly deviates from mobility patterns of random walkers. Mean first passage times and transition counts relate via a power-law with slope -1/3. We compare our results with studies where human mobility was measured via mobile phone data and find striking similarities. | Medizinische Universität, Bauteil 88, Seminarraum 513, Ebene 3 |

2010-12-03 14:30-16:00 | P. Klimek (Medical University of Vienna) | An Evolutionary Economics Model for Innovation Processes | |

2010-11-19 11:00-12:00 | M. D'Errico (Medical University of Vienna) | The economy of the game Pardus | Seminarraum BT 86 |

2010-11-19 14:30-16:00 | H. Rumpf (Universität Wien) | Is Gravity an Entropic Force? | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5. Stock |

2010-10-29 11:00-12:30 | V. Traag (Universite Catholique de Louvain) | Cooperation, Reputation & Gossiping Abstract Explaining the breadth of human cooperation is a prime challenge in both the social sciences and biology. One possible mechanism focuses on the role of reputation in cooperation. Usually it is assumed that such a reputation is objective--that is, the same for all agents. We develop a model where reputations are private and synchronized through the sharing of information, i.e. gossiping. Interpreting cooperation between two agents as a positive link and defection as a negative link, this model shows an interesting connection to social balance theory. Furthermore, we study its evolutionary stability. | Medizinische Universität, Bauteil 88, Bibliothek 'Zentrum für Medizinische Statistik, Informatik und Intelligente Systeme', Ebene 3 |

2010-10-22 14:30-16:00 | F. Verstraete (Universität Wien) | Renyi entropies and Hypothesis testing for quantum systems | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2010-06-25 14:30-16:00 | S. Thurner (Medical University of Vienna) | Lecture Series Entropy 2010: Classification of complex statistical systems in terms of stability and a thermodynamical derivation of their entropy and distribution functions Abstract Strongly interacting statistical systems – complex systems in particular – can change their macro-
scopic properties merely as a function of the number of their constituents. Examples include neurons,
state-forming insects, financial markets, etc. where systemic properties of small systems can differ
drastically from those of a large system built from the same components. The origin of this property
is not understood on fundamental grounds. Here we explore this phenomenon from first principles
within a thermodynamical framework, by asking about the consequences of bringing interacting sub-
systems in thermal contact, where the first three Kinchin axioms hold but the 4th is violated. We
show that all sufficiently interacting statistical systems fall into two categories: systems which are
asymptotically stable, and those which are asymptotically unstable, meaning that small changes in
system size can lead to a drastic increase in entropy. We argue that complex systems belong to this
unstable class which make drastic qualitative changes possible as a function of system size. Under
the same conditions we then derive the unique asymptotic entropy, Scd = Gamma (d + 1,1 - c ln pi)
(c, d constants) which covers all equivalence classes of asymptotically stable and unstable, i.e. all
interacting and non-interacting systems. The corresponding distribution functions are special forms
of Lambert-W exponentials. As special cases they contain Boltzmann, stretched exponential and
Tsallis distributions (power-laws) – all widely abundant in nature. This is, to our knowledge, the
first ab initio justification for the existence of generalized entropies. | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5. Stock |

2010-06-24 | P. Klimek (Medical University of Vienna) | Defensio | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2010-06-23 16:00-17:30 | C. Brukner (Universität Wien) | Lecture Series Entropy 2010: Information Complementary Relation in Quantum Mechanics | Josef Stefan Hörsaal, Boltzmanngasse 5, A-1090 Vienna, 3rd floor |

2010-06-18 14:30-16:00 | C. Tsallis (CBPF Rio de Janeiro and Santa Fe Institute) | Lecture Series Entropy 2010: Statistical Mechanics of Systems lying outside the domain of validity of the Boltzmann-Gibbs theory Abstract The celebrated statistical mechanics introduced by Boltzmann and Gibbs more than one century ago lie (for classical systems, for instance) on hypotheses such as ergodicity and mixing. Strongly chaotic systems, with positive maximum Lyapunov exponent, satisfy requirements of this sort. Within this realm, relevant random variables are probabilistically independent or nearly so. It is for such situations, and related quantum ones, that the central limit theorem and the standard entropy (Boltzmann, Gibbs, von Neumann, Shannon) exhibit their well known utility and connections with classical thermodynamics. What can be done outside this world? Can we approach such anomalous, and nevertheless ubiquitous, cases on thermostatistical grounds similar to the usual ones? For wide classes of such systems the answer appears to be positive, by appropriately generalizing the entropy and, consistently, the central limit theorem. Some central concepts as well as typical verifications and applications for natural, artificial and social systems will be briefly presented. BIBLIOGRAPHY: (i) C. Tsallis, Entropy, in Encyclopedia of Complexity and Systems Science (Springer, Berlin, 2009); (ii) C. Tsallis, Introduction to Nonextensive Statistical Mechanics - Approaching a Complex World (Springer, New York, 2009); (iii) S. Umarov, C. Tsallis, M. Gell-Mann and S. Steinberg, J. Math. Phys. 51, 033502 (2010); (iv) CMS Collaboration, J. High Energy Phys. 02, 041 (2010); (v) http://tsallis.cat.cbpf.br/biblio.htm | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5. Stock |

2010-06-11 14:30-16:00 | E. Ortega (Universität Wien) | Lecture Series Entropy 2010: Entropy in Many Body Quantum Systems | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5. Stock |

2010-06-09 16:00-17:30 | J. Yngvason (Universität Wien) | Lecture Series Entropy 2010: The Entropy of Classical Thermodynamics | Josef Stefan Hörsaal, Boltzmanngasse 5, A-1090 Vienna, 3rd floor |

2010-05-18 14:30-16:00 | R. Hanel (Medical University of Vienna) | A simple model of Evolutionary dynamics: Minimally nonlinear systems as model of genetic regulatory network dynamics. | Medizinische Universität, Bauteil 86, Ebene 3 |

2010-05-14 14:00-15:30 | M. Poechacker (Medical University of Vienna) | Stability Analysis of Random Catalytic Network Dynamics in a Linear Model of Gene-Regulation | Medizinische Universität, Bauteil 86, Ebene 3 |

2010-05-07 15:15-16:00 | M. Szell (Medical University of Vienna) | A Multiplex View of Organization and Social Balance in Large-scale Social Networks Abstract Human societies can be regarded as large numbers of locally interacting agents, connected by a broad range of social and economic relationships. These relational ties represent e.g. the feeling a person has for another (friendship, enmity, love), communication, exchange of goods, or behavioral interactions. Each type of relation spans a social network on its own. A whole society can be understood systemically only by uncovering interactions between different networks. Here we analyze a complete, multi-relational social network of a society consiting of over 300,000 players of a massive multiplayer online game. We extract networks of six different types of one-to-one interactions between the players. Three of them carry a positive connotation (friendship, communication, trade), three a negative (enmity, armed aggression, punishment). We first analyze these types of networks as separate entities and find that negative interactions differ from positive interactions by their lower reciprocity, weaker clustering and fatter-tail degree distribution. We then proceed to explore how the inter-dependence of different network types determines the organization of the social system. In particular we study correlations and overlap between different types of links and demonstrate the tendency of individuals to play different roles in different networks. As a demonstration of the power of the approach we present the first empirical large-scale verification of the long-standing structural balance theory, by focusing on the specific multiplex network of friendship and enmity relations. | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2010-05-07 14:30-15:15 | M. Göll (Universität Wien) | Lattice Models as Algebraic Dynamical systems, part II | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5. Stock |

2010-04-23 14:30-16:00 | C. Losert-Valiente Kroon (Erwin Schrödinger International Institute for Mathematical Physics) | Stochastic Population Dynamics in Astrochemistry and Aerosol Science | Medizinische Universität, Bauteil 88, Bibliothek 'Zentrum für Medizinische Statistik, Informatik und Intelligente Systeme', Ebene 3 |

2010-04-16 14:30-16:00 | M. Göll (Universität Wien) | Lattice models as algebraic dynamical systems | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2010-03-26 14:30-16:00 | R. Hanel (Medical University of Vienna) | Minimally non-linear systems: Toy genetic regulatory network dynamics at the edge of chaos | |

2010-03-05 14:30-16:00 | P. Schreivogl (Universität Wien) | Topological Strings and the Melting Crystal | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2010-01-29 14:30-16:00 | H. Hüffel (Universität Wien) | From Classical Statistical Physics to Quantum Mechanics: Dimers and Crystal Melting | |

2010-01-15 14:30-16:00 | M. Sadilek (Universität Wien) | Quantenaspekte aktiver Bewegung und anharmonische Effekte in Kristallen | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2010-01-08 14:30-16:00 | P. Klimek (Medical University of Vienna) | Evolutionary Dynamics | Medizinische Universität, Bauteil 86, Ebene 2 |

2009-12-11 14:00-15:30 | H. Meyer-Ortmanns (Jacobs University, Bremen) | Stochastic Transport Processes on Networks | Medizinische Universität, Bauteil 86, Ebene 2 |

2009-12-04 14:30-16:00 | A. Glück (Universität Wien) | Swarms with canonical active Brownian motion | Erwin Schrödinger Hörsaal, Boltzmanngasse 5, 5.Stock |

2009-11-20 14:30-16:00 | M. Szell (Medical University of Vienna) | Measuring social dynamics in a massive multiplayer online game Abstract Quantification of human group-behavior has so far defied an empirical, falsifiable approach. This is due to tremendous difficulties in data acquisition of social systems. Massive multiplayer online games (MMOG) provide a fascinating new way of observing hundreds of thousands of simultaneously socially interacting individuals engaged in virtual economic activities. We have compiled a data set consisting of practically all actions of all players over a period of three years from a MMOG played by 300,000 people. This large-scale data set of a socio-economic unit contains all social and economic data from a single and coherent source. Players have to generate a virtual income through economic activities to 'survive' and are typically engaged in a multitude of social activities offered within the game. Our analysis of high-frequency log files focuses on three types of social networks, and tests a series of social-dynamics hypotheses. In particular we study the structure and dynamics of friend-, enemy- and communication networks. We find striking differences in topological structure between positive (friend) and negative (enemy) tie networks. All networks confirm the recently observed phenomenon of network densification. We propose two approximate social laws in communication networks, the first expressing betweenness centrality as the inverse square of the overlap, the second relating communication strength to the cube of the overlap. These empirical laws provide strong quantitative evidence for the Weak ties hypothesis of Granovetter. Further, the analysis of triad significance profiles validates well-established assertions from social balance theory. We find overrepresentation (underrepresentation) of complete (incomplete) triads in networks of positive ties, and vice versa for networks of negative ties. Empirical transition probabilities between triad classes provide evidence for triadic closure with extraordinarily high precision. For the first time we provide empirical results for large-scale networks of negative social ties. Whenever possible we compare our findings with data from non-virtual human groups and provide further evidence that online game communities serve as a valid model for a wide class of human societies. With this setup we demonstrate the feasibility for establishing a 'socio-economic laboratory' which allows to operate at levels of precision approaching those of the natural sciences. | Medizinische Universität, Bauteil 86, Ebene 2 |

2009-09-10 11:00-12:30 | R. Sinatra (University of Catania) | Networks of recurrent motifs from sequences of symbols Abstract We present a general method to convert an ensemble of symbolic sequences into a weighted directed network. The nodes of the network are short motifs of the ensemble, while the directed links and their weights are defined from statistically significant co-occurrences of two motifs in the same sequence. The method is shown to be able in correlating sequences with functions in protein data, and might find useful applications for structure discovery as well as in computational linguistic and in the theory of dynamical systems. | Medizinische Universität, Bauteil 86, Ebene 2 |

Seminars before 2009-09 are not available.