ABSTRACT. The evolution of life in our biosphere has been marked by several major innovations. Such major complexity shifts include the origin of cells, genetic codes, or multicellularity to the emergence of language, cognition, or even consciousness. Understanding the nature and conditions for evolutionary innovation is a major challenge for evolutionary biology. Along with data analysis, phylogenetic studies and dedicatedexperimental work, theoretical and computational studies are an essential part of this exploration. With the rise of synthetic biology, evolutionary robotics, artificial life, and advanced simulations, novel perspectives to these problems have led to an emerging new synthesis, where evolutionary innovations can be understood in terms of phase transitions, as defined in physics. Such mapping (if correct) would help in defining a general framework to establish a theory of evolutionary change and the role played by chance and constraints in shaping living complexity.

ABSTRACT. Network controllability and its robustness has been widely studied. However, analytical methods to calculate network controllability with respect to node removals are currently lacking. This paper develops methods, based upon generating functions for the in- and out-degree distributions, to approximate the minimum number of driver nodes needed to control directed networks, during random and targeted node removals. By validating the proposed methods on synthetic and real-world networks, we show that our methods work very well in the case of random node removals and reasonably well in the case of targeted node removals, in particular for moderate fractions of attacked nodes.

ABSTRACT. Real-world infrastructures are often subject to random failures or intentional attacks that can significantly impact their robustness and hence the many processes taking place on them. In this paper, we focus on the robustness of complex networks by proposing a link addition strategy for improving network robustness. The approach exploits the incremental computation of the Moore-Penrose pseudoinverse matrix to efficiently compute the effective graph resistance when a new link is added to the network. Experiments on both real-world and synthetic data sets show that the strategy provides a good trade-off between a low percentage error of effective graph resistance with respect to the exhaustive search and the simulation time needed to obtain the optimal link.

ABSTRACT. Multi-layer Complex networks are commonly used for modeling and analyzing biological entities. This paper presents a new computational framework called COMBO (Combining Multi Bio Omics) for generating and analyzing heterogeneous multi-layer networks. Our model uses gene expression and DNA-methylation data. The power of COMBO relies on its ability to join different omics to study the complex interplay between various components in the disease. We tested the reliability and versatility of COMBO on colon and lung adenocarcinoma cancer data obtained from the TCGA database.

ABSTRACT. This paper compares network information against chemical information for the problem of drug interaction prediction. Drug interactions can be studied as a network, with the drugs represented as nodes, and interactions as edges. There is also the additional information of the chemical formula of each drug. In particular, we explore various topological features of the network and the effect they have on network based and chemical based link prediction. Node embeddings and chemical formula embeddings are used to train classifiers. Then the results are compared based on degree, betweeness, and core centralities. The experimental evaluation is performed on a data set from DrugBank.

ABSTRACT. The big break in data collection tools of large-scale networks from biological, social, and technological domains expands the challenge of their visualization and processing. Numerous structural and statistical backbone extraction techniques aim to reduce the network’s size while preserving its gist. Here, we perform an experimental comparison of seven main statistical methods in an air transportation case study. Correlations analysis shows that Marginal Likelihood Filter (MLF), Locally Adaptive Network Sparsification Filter (LANS), and Disparity Filter are biased toward high weighted edges. We compare the extracted backbones us- ing four indicators: the size of the largest component, the number of nodes, edges, and the total weight. Results show that techniques based on a binomial distribution null model (MLF and Noise Corrected Filter) tend to retain many edges. Conversely, Disparity Filter, Polya Urn Fil- ter, LANS Filter, and Global Statistical Significance Filter (GLOSS) are pretty aggressive in filtering edges. The ECM Filter lies between these two behaviors. These results may guide users in selecting appropriate techniques for their applications

ABSTRACT. The Mantel Test, discovered in the 1960s, determines whether two distance metrics on a graph are related. We describe DC-RST, an algorithm to accelerate a key step of a network science statistical computation associated with DimeCost, an approach that is faster the Mantel Test. DC-RST is a parallel, divide-and-conquer algorithm to compute a random spanning tree of a complete graph on $n$ vertices. Relative to an implementation of Wilson's sequential random-walk algorithm, on a system with 48 cores, DC-RST was up to 4X faster when first creating random partitions and up to 20X faster without this sub-step. DC-RST is shown to be a suitable replacement for Wilson's sequential algorithm through a combination of theoretical and statistical results.

ABSTRACT. We study the problem of recovering a planted hierarchy of partitions in a network. The detectability of a single planted partition has previously been analysed in detail and a phase transition has been identified below which the partition cannot be detected. Here we show that, in the hierarchical setting, there exist additional phases in which the presence of multiple consistent partitions can either help or hinder detection. Accordingly, the detectability limit for non-hierarchical partitions typically provides insufficient information about the detectability of the complete hierarchical structure, as we highlight with several constructive examples.

ABSTRACT. Methods for consensus partitioning are relevant to the network research, as pointed out by Lancichinetti and Fortunato (2012) and Liu et al. (2022). This paper reports of theoretical and computational results related to an original concept of consensus clustering involving what we call the projective distance between partitions. This distance is defined as the squared difference between a partition incidence matrix and its image over the orthogonal projection in the linear space spanning the other partition incidence matrix.

ABSTRACT. Here we present a method to decluster the earthquake catalogue by network analysis. Community detection is one of the most progressive fields in complex network analysis, due to its promising value in practical applications. Earthquake networks have communities consisting of main events and related events. We aim to separate these communities using known and advanced community detection techniques, which will serve our purpose of declustering.

ABSTRACT. In complex networks, network dismantling aims at finding an optimal set of nodes (or edges) such that the removal of the set from the network will lead to the disintegration of the network, that is, the size of the giant/largest connected component is not bigger than a specific threshold (for instance, $1\%$ of the original network size). The existing algorithms addressed this topic can be divided into two closely related but different categories: vertex separator-oriented algorithms and edge separator-oriented algorithms. There has been a lot of research on these two categories, respectively. However, to the best of our knowledge, less attention has been paid to the relation between the vertex separator and edge separator. In this paper, we studied the separator transformation (ST) problem between the separator of the vertexes and edges. We approximated the transformation from edge separator to vertex separator using Vertex Cover algorithm, while approximated the transformation from vertex separator to edge separator using an Explosive Percolation (EP) approach. Moreover, we further analyzed the results of the vertex-edge separator transformation through the explosive percolation method in detail. The transformation problem in network dismantling opens up a new direction for understanding the role of the vital nodes set and edges set as well as the vulnerability of complex systems.

ABSTRACT. Most models of information diffusion online rely on the assumption that pieces of information spread independently from each other. However, several works pointed out the necessity of investigating the role of interactions in real-world processes, and highlighted possible difficulties in doing so: interactions are sparse and brief. As an answer, recent advances developed models to account for interactions in underlying publication dynamics. In this article, we propose to extend and apply one such model to determine whether interactions between news headlines on Reddit play a significant role in their underlying publication mechanisms. After conducting an in-depth case study on 100,000 news headline from 2019, we retrieve state-of-the-art conclusions about interactions and conclude that they play a minor role in this dataset.

ABSTRACT. The epistemic position of an agent often depends on their position in a larger network of other agents who provide them with information. In general, agents are better off if they have diverse and independent sources. Sullivan et al. (2020) developed a method for quantitatively characterizing the epistemic position of individuals in a network that takes into account both diversity and independence; and presented a proof-of-concept, closed-source implementation on a small graph derived from Twitter data (Sullivan et al., 2020). This paper reports on an open-source reimplementation of their algorithm in Python, optimized to be usable on much larger networks. In addition to the algorithm and package, we also show the ability to scale up our package to large synthetic social network graph profiling, and finally demonstrate its utility in analyzing real-world empirical evidence of 'echo chambers' on online social media, as well as evidence of interdisciplinary diversity in an academic communications network.

ABSTRACT. In this research we explore how social platforms user express about COVID-19, more specifically we searched for tweets related to COVID-19 vaccine hesitancy in spanish speaking Twitter. To do this, we gathered tweets with the keyterms "vaccine", "sputnik v", "Moderna", "Pfizer", among others. Then, we identified hashtags of interest related to these topics, those than would express ideas of hesitancy, like "#IDontVaccine", "#Plandemic", "Covid1984", to name a few. We collected 57,212 tweets (retweets not included), then we performed a series of analyses with the aim of describe these coversations. NLP techniques, sentiment analysis, text analysis and communities detection were the tools we used to help us characterize the opinions the users expressed.

ABSTRACT. We investigate the case in which each piece of information has a numerical proxy representing its quality, and the higher the quality, the greater are the chances of being transmitted further in the network. The model allows us to study how sorting information in the agent's attention list according to their quality, node's influence and popularity affect the overall system's quality, diversity and discriminative power. We compare the three scenarios with a baseline model where the information is organized in a first-in first-out manner. Our results indicate that such an approach intensifies the exposure of high-quality information increasing the overall system's quality while preserving its diversity. However, it significantly decreases the system's discriminative power.

ABSTRACT. In the last decades, social media have opened novel way to participate to the social life. This has unfolded disinformation proliferation on many topics. Malicious actors use social media platforms to covertly influence the choices, actions, and interactions of individual users. The strategies of malicious actors are informed by their knowledge of the demographic characteristics of users and how those characteristics can be targeted to disrupt democratic processes. Therefore, to increase individuals’ and societies’ resilience in the face of targeted disinformation and malicious attacks, regulatory approaches need to address the vulnerabilities of individual users as well as ensuring they have appropriate safety systems to mitigate these abuses. The work we present indicates that social media users differently react to real life events exhibiting a diverse set of responses, ranging from moderate to high levels of hateful content shared in online discussions, which is inevitably linked to the spread of disinformation. This opens to the possibility of classifying the users based on their susceptibility to offline events and online harms, and of adapting the flow of information they are exposed to by designing and evaluating a set of per-class and/or personalized intervention strategies, including the demotion of disinformation and hyper-partisan content.

ABSTRACT. In April 2022, the French presidential election took place, and social media played a prominent role in it. By analyzing more than 150 million interactions on French Twitter, this study aims to provide evidence of coordinated behaviors from political parties. We find that extreme parties, left and right, appear with a particular internal structure compared to moderate parties. Moreover, by examining similar patterns in community structures but also in duplicated tweets, we unveil online astroturfing strategies of the main parties online, and in particular the extreme right.

ABSTRACT. Most real-world networks come as evolving topologies whose nodes and edges appear/disappear as time goes by. Accordingly, the meso-scale substructures hidden in such systems are subject to constant evolution as well. The task of Dynamic Community Detection represents the challenge of identifying evolving groups in dynamic networks and track group mutations over time. Several categories of evolutionary events that characterize the life-cycle of a community have been proposed in the literature. However, all these mentioned events address topological changes only. Moreover, they never mention the importance of node metadata, widely used nowadays to identify both well-connected and homogeneous communities. Since more and more algorithms begin to be applied to attributed dynamic networks we propose in this preliminary work to reason about a categorization of attribute-aware community events.

ABSTRACT. As COVID-19 began to sweep the world in 2019, the mobility of more than 4 billion people worldwide in 2020 was severely limited, which severely affected people's life. From China's perspective, due to the vast territory and different developments of provincial-level regions, the COVID-19 transmission pattern is different to varying degrees. Therefore, we studied the spread of the epidemic with the air transport network within the framework of the urban development network. In the urban development network, cities are taken as nodes and airlines as links, and the comprehensive indicator composed of political status, economic strength, city size, and population composition is considered. On this basis, the spatio-temporal framework of the air transport network with airports as nodes and flights as links is constructed, and factors such as social distance, population mobility density, and infection rate are considered. In the double-layer network of urban development network--air transport network, the urban development network not only decides the degree centrality, closeness centrality, betweenness, and eigenvector centrality of the air transport network. In addition, the time delay in responding to COVID-19, the length of time the aircraft was grounded, and the number of air routes closed are all related to the previous flow between airport pairs and the comprehensive indicator of the urban node pairs. All of these factors are also directly related to local mortality and cure rates due to COVID-19. Only by grasping the spread of the epidemic in various regions of China can we take precise prevention and control measures according to local conditions to ensure the safety of air transport.

ABSTRACT. Generally, in biological and ecological disciplines, the concept of adaptation references ecosystems changing as system perturbations arise or species entering a niche left after a sudden climatic shift. This paper extends these concepts to road transportation networks and focuses on the adaptability of commuters to traffic disruptions. Specifically, we apply an Ecosystem Network Analysis (ENA) and information-theoretic framework to demonstrate the potential, calculated as the number of alternate options available throughout a given network, for adaptive routing optimization on road networks. An initial assessment of balanced metrics of resilience and efficiency, calculated from 13 Metropolitan Statistical Area (MSA) road networks, is performed. These metrics are then compared with their respective commuter delay levels which indicates a correlation between the balance of resilient and efficient networks and the annual commuter delay of each MSA. Whereas road network topologies that demonstrate either a highly efficient or highly resilient network structure show a tendency for higher commuter delay levels, road networks that balance efficiency and resiliency suggest a tendency for lower commuter delay levels.

ABSTRACT. Interpersonal synchrony is associated with better interpersonal affiliation. No matter how well-affiliated people are, interruptions or transitions in synchrony rebound to occur. One might intuitively expect that transitions in synchrony negatively affect affiliation or liking. Empirical evidence, however, suggests that time periods with interruptions in synchrony may favor affiliation or liking even more than time periods without interruptions in synchrony. This paper introduces a controlled adaptive network model to explain how persons’ affiliation might benefit from transitions in synchrony over and above mean levels of synchrony. The adaptive network model was evaluated in a series of simulation experiments for two persons with a setup in which a number of scenarios were encountered in different (time) episodes. Our controlled adaptive network model may serve as a foundation for more realistic virtual agents with regard to synchrony transitions and their role in affiliation.

ABSTRACT. The \textbf{A}rtificial \textbf{B}enchmark for \textbf{C}ommunity \textbf{D}etection graph (\ABCD) is a random graph model with community structure and power-law distribution for both degrees and community sizes. The model generates graphs with similar properties as the well-known \textbf{LFR} one, and its main parameter $\xi$ can be tuned to mimic its counterpart in the \textbf{LFR} model, the mixing parameter $\mu$.

In this paper, we investigate various theoretical asymptotic properties of the \ABCD\ model. In particular, we analyze the modularity function, arguably, the most important graph property of networks in the context of community detection. Indeed, the modularity function is often used to measure the presence of community structure in networks. It is also used as a quality function in many community detection algorithms, including the widely used \emph{Louvain} algorithm.

ABSTRACT. Many complex systems involve direct interactions among more than two entities and can be represented by hypergraphs, in which hyperedges encode higher-order interactions among an arbitrary number of nodes. To analyze structures and dynamics of given hypergraphs, a solid practice is to compare them with those for randomized hypergraphs that preserve some specific properties of the original hypergraphs. Here we introduce a family of such reference models for hypergraphs, called the hyper dK-series, that we proposed in Ref. [1]. The hyper dK-series preserves up to the individual node’s degree, node’s degree correlation, node’s so-called redundancy coefficient, and/or the hyperedge’s size, depending on the values of the parameters d_v and d_e. Furthermore, we numerically find that higher-order hyper dK-series more accurately preserves the shortest path length of the original hypergraph, which the method does not intend to preserve. We also apply the hyper dK-series to numerical simulations of evolutionary game dynamics on an empirical social hypergraph. We find that the hyperedge’s size affects evolutionary dynamics more than any of the node’s properties and that the node’s degree correlation and redundancy in the empirical hypergraphs promote cooperation.

ABSTRACT. The maximum entropy production principle is a generalization of the maximum entropy principle which holds in many areas of science, from physics, through chemistry to biology. It states that not only equilibrium systems produce maximum entropy, but also out of equilibrium ones. It has been applied to many situations, but mostly from the point of view of thermodynamic entropy. This work presents an information theoretic version of this principle as an approach to modelling dynamic random processes. A general evolution equation for a probability distribution is deduced assuming the maximum entropy production principle and a general set of constraints. The equation is then used to predict the evolution of an undirected binary network from an initial condition into an asymptotic condition. We will see that the equilibrium state corresponds to the one predicted by the maximum entropy principle, and that the evolution is comparable to predicitons from other methods.

ABSTRACT. Random graphs play a central role in network analysis. The Chung-Lu random graph model is one particularly popular model which connects nodes according to their desired degrees to form a specific degree distribution in expectation. Despite its popularity, the standard Chung-Lu graph generation algorithms are susceptible to significant distribution errors when generating simple graphs. In this manuscript, we suggest multiple methods for improving the accuracy of Chung-Lu graph generation by computing node weights which better recreate the desired output degree distribution. We show that each of our solutions offer a significant improvement in distribution accuracy.

ABSTRACT. Complex networks encountered in biology, ecology, sociology and technology often contract due to node failures, infections or attacks. The ultimate failure, taking place when the network fragments into disconnected components was studied extensively using percolation theory. We show that long before reaching fragmentation, contracting networks lose their distinctive features. In particular, we identify that a very large class of network structures, which experience a broad class of node deletion processes, exhibit a stable flow towards universal fixed points, representing a maximum-entropy ensemble, which in the pure contraction scenario is the Erdos-Renyi ensemble. Under more general combination of growth and deletion processes the resulting degree distribution is Poisson-like. This is in sharp contrast to network growth processes that often lead to scale-free networks. We provide a full phase diagram of the different structures that occur in such processes, and identify a structural and a dynamical phase transition. These results imply that contracting networks in the late stages of node failure cascades, attacks and epidemics reach a common structure, providing a unifying framework for their analysis.

ABSTRACT. Hyperbolic networks models has gained much attention recently, since they can naturally explain the scale-free, highly clustered and small-world properties of real networks. Here we show for the popularity-similarity optimization model that the generated networks become also extremely modular in the large network size limit, although there is no explicit community formation mechanism in the model definition. According to obtained results supported by numerical simulations, when the network size is increased, the modularity can get arbitrarily close to - and more than that, can even approach - one.

ABSTRACT. In hyperbolic geometric networks the vertices are embedded in a latent metric space and the edge probability depends on the hyperbolic distance between the nodes. These models allows to produce networks with high clustering and scale-free degree distribution, where the coordinates of the vertices abstract their centrality and similarity. By building on the principles of hyperbolic models, in this paper we introduce the Hyperbolic Geometric Block Model, which allows to obtain high-clustering and scale-free networks while preserving the desired group mixing structure. We additionally study a parametric network model whose edge probability depends on both the distance in an explicit euclidean space and the distance in a latent geometric space. Through extensive simulations on a stylized city of 10K inhabitants, we provide experimental evidence of the robustness of the HGBM model and of the possibility to combine a latent and an explicit geometry to produce data-driven social networks that exhibit all the main features observed in empirical networks.

ABSTRACT. Complex network characteristics such as centralities have lately started to be associated with passenger flows at metro stations. Centralities can be leveraged in an effort to develop fast and cost-efficient passenger flow predictive models. However, the accuracy of such models is still under question, and the most appropriate predictors are yet to be found. In this sense, this study attempts to investigate appropriate predictors, and develop a predictive model for daily passenger flows at metro stations, based exclusively on spatial attributes. Using the Athens metro network as a case study, a linear regression model is developed, with node degree, betweenness and closeness centralities of the physical network, node strength of the substitute network, and a dummy variable of station importance being as covariates. An econometric analysis validates that a linear model is suitable for associating centralities with passenger flows, while model’s evaluation metrics indicate satisfying accuracy. In addition, a machine learning benchmark model is utilized to further investigate variable significance and validate the accuracy of the linear model. Last but not least, both models are utilized for predicting passenger flows at the new metro stations of the Athens metro network expan-sion. Findings suggest that node strength of the substitute network is a powerful predictor and the most significant covariate of both models; both models’ accuracy and predictions converge to a great extent. The model developed is expected to facilitate medium-term disruption management through providing information about metro passenger flows at low cost and high speed.

ABSTRACT. We study the emergence of congestion patterns in urban networks by modeling vehicular interaction by means of a simple traffic rule and by using a set of measures inspired by the standard Betweenness Centrality (BC). We consider a topologically heterogeneous group of cities and simulate the network loading during the morning peak-hour by increasing the number of circulating vehicles. At departure, vehicles are aware of the network state and choose paths with optimal traversal time. Each added path modifies the vehicular density and travel times for the following vehicles. Starting from an empty network and adding traffic until transportation collapses, provides a framework to study network's transition to congestion and how connectivity is progressively disrupted as the fraction of impossible paths becomes abruptly dominant. We use standard BC to probe into the instantaneous out-of-equilibrium network state for a range of traffic levels and show how this measure may be improved to build a better proxy for cumulative road usage during peak-hours. We define a novel dynamical measure to estimate cumulative road usage and the associated total time spent over the edges by the population of drivers. We also study how congestion starts with dysfunctional edges scattered over the network, then organizes itself into relatively small, but disruptive clusters.

ABSTRACT. Cities are complex systems which process information, evolve, and adapt to their environment. Understanding urban dynamics at individual level -- but also as the outcome of collective human behaviour -- will open the doors to uncountable applications ranging from enhancing the sustainability and the resilience of the city to improving health and well-being of its inhabitants. To understand how complex systems -- and cities more specifically -- operate, it is thus important to quantify how information is processed in terms of integration and segregation.

The goal of our paper is indeed to better characterize the functional organization of a city through the lens of network science. To this aim we measure to which extent different areas of the city facilitate human flows -- i.e., functional integration -- and to which extent there are separate clusters of areas characterized by within-cluster flows larger than between-cluster flows -- i.e., functional segregation. By considering those measures simultaneously, it is possible to characterize how well human flows mix through the city according to the existing distribution of venues and the way residents use them. In fact, the dichotomy between integration and segregation -- often improperly used as antonyms -- is relevant for improving our understanding of the interplay between urban structure and human behavior.

A particularly relevant perspective is provided by activity-aware information such as the one provided by users of Foursquare -- a leading location intelligence platform -- which allows people to investigate human flows at different scales with unprecedented detail. This type of data is of special interest because one can investigate the interplay between the structure of a city and the dynamics of its inhabitants to gain novel insights about the functional organization of the underlying urban ecosystem. In this work, we stratify human activities in Foursquare to build network models describing the human movements across the urban space -- from hours to months -- within the different areas of 10 different metropolitan systems worldwide.

The metadata of the venues include a category field which describes the type of venue in great detail. We defined a set of macro-categories we used to define a limited number of layers). By disentangling the mobility flows into a multilayer network structure, we are able to quantify the differences in the functional organization of the different ``cities within a city'' that are outlined by movement between different types of activities in a limited number of layers. The multilayer analysis allows to highlight the strikingly distinct views on the functional organization of a city extracted by isolating intra- or inter-layer flows. These maps outline the different ``cities within the city'' which we disentangle by decoupling the urban flows into activity-aware multilayer networks.

By comparing the results obtained from empirical networks with those obtained for synthetic networks of different sizes (e.g, Random Geometric Networks and Watts-Strogatz small world), we discover that many features of complex megacities can be understood from simple mechanisms related to geometric constraints and city's characteristic size, with larger cities tending to be more segregated and less integrated. Random geometric models with long-range connections seem to be a good candidate to reproduce the most salient features measured from empirical data, and further research is required in this direction to confirm this finding for a wider spectrum of urban systems. Interestingly, the interplay between heterogeneities in the underlying network connectivity and spatial constraints might be responsible for the emergence of integrated/segregated structures that might be reflected in the functional organization of the city, and future research should point in this direction to gain new insights.

ABSTRACT. How do pedestrians choose their paths within city street networks? Researchers have tried to shed light on this matter through strictly controlled experiments, but an ultimate answer based on real-world mobility data is still lacking. Here, we analyze salient features of human path planning through a statistical analysis of a massive dataset of GPS traces, which reveals that (1) people increasingly deviate from the shortest path when the distance between origin and destination increases and (2) chosen paths are statistically different when origin and destination are swapped. We posit that direction to goal is a main driver of path planning and develop a vector-based navigation model; the resulting trajectories, which we have termed pointiest paths, are a statistically better predictor of human paths than a model based on minimizing distance with stochastic effects. Our findings generalize across two major US cities with different street networks, hinting to the fact that vector-based navigation might be a universal property of human path planning.

ABSTRACT. The intertwining of socio-spatial complexity with that of price formation leads to challenging questions in spatial statistics when modelling urban real estate markets and their dynamics. The exact same apartment typically will not have the same value depending on its location in the city – due to the specifics of the local neighborhood and other qualitative features. Traditional methods rely on so-called hedonic prices and implicit markets. These are essentially regression methods, that may be modified in order to incorporate spatial effects via geographically weighted regressions (GWR). Here we present an altogether new paradigm, using a complex network approach: the city itself is thought of as a network of locations that functions as a neural network, "learning" its housing market. We apply this paradigm to data from the Paris housing market between 2016 and 2021 and show that this approach outperforms GWR.

ABSTRACT. We present our current work aiming to find a solution for using drones in parcel delivering over cities, looking for a balance between the legitimate conflicting stakeholders demands. We present our droneways proposal as an enabling technology that allow controlled use of urban air space in order to make feasible and accountable the use of these aerial vehicles with minimum impact for the people and environment. In addition, we show the performance of two heuristics to find efficient subgraphs on geographical complex networks, applying them to real cities. This work provides contributions for smart cities planning, drones usage, and business modeling.

ABSTRACT. This work introduces a communication protocol that allows to reach multiconsensus in a multi-agent system by exploiting the symmetries of the interaction network. In particular, this protocol ensures the convergence to a solution that is parallel to the leading eigenvector of the network adjacency matrix, so that symmetric nodes reach the same consensus value. In addition to the case of perfect symmetries, we have analyzed the system dynamics in the presence of approximate symmetries and we have provided an estimate of the difference between the final state reached by the system in the two cases.

ABSTRACT. In this work, we propose a multilayer control protocol for synchronization of network dynamical systems under limited resources. In addition to the layer where the interactions of the system takes place, i.e., the backbone network, we propose a second, adaptive layer, where the edges are added or removed according to the edge snapping mechanism. Different from classic edge snapping, the inputs to the edge dynamics are modified to cap the number of edges that can be activated. After studying local stability of the overall network dynamics, we illustrate the effectiveness of the approach on a network of Rössler oscillators and on the Italian high-voltage power grid model.

ABSTRACT. In a globalizing world, the shareholder may have substantial influence. This power is amplyfied by the global ownership network because they can indirectly control comapnies and collect dispersed voting rights. Previous works found the government of China is the most influential shareholder in 2016, creating a hierarchical ownership structure. However, they only focus on ultimate owners and dismiss the major financial institutions especially in the United States because some of them are not ultimate owners. This paper aims to assess the contribution of intermediate companies to the indirect control of shareholders.

The Shapley-Shubik power index is known as a canonical model in the literature of measuring voting power in a weighted majority vote but it considers only direct control. Extending it to an ownership network, Network Power Index (NPI) is proposed as the measurement of indirect control of shareholders. While NPI reveals the distribution of indirect power of ultimate owners by taking into account the transitivity and consolidation of fragmented vorting rights, it ignores the crucial aspect of control structure in a network because the power of ultimate owners stems from intermediate companies which connect upstream and downstream companies. To overcome this limittion, we propose a model to identify the channels through which the ultimate owners' power of corporate control over companies travel through the global shareholding network.

ABSTRACT. We report on the unusual dynamical features of the transients from random initial conditions to a fully synchronized (one-cluster) state, which are observed in numerical simulations with the Kuramoto model and experiments with oscillatory nickel electrodissolution. In particular, the numerical simulations revealed that certain networks (e.g., globally coupled or dense Erdos-Renyi random networks) showed relatively simple behavior with monotonic increase of the Kuramoto order parameter from the random initial condition to the fully synchronized state and that the transient times exhibited a unimodal distribution. However, some modular networks with bridge elements were identified which exhibited non-monotonic variation of the order parameter with local maximum and/or minimum. In these networks, the histogram of the transients times became bimodal and the mean transient time scaled well with inverse of the magnitude of the second largest eigenvalue of the network Laplacian matrix. The non-monotonic transients increase the relative standard deviations from about 0.3 to 0.5, i.e., the transient times became more diverse. The non-monotonic transients are related to generation of phase patterns where the modules are synchronized but approximately anti-phase to each other. The predictions of the numerical simulations were demonstrated in a population of coupled oscillatory electrochemical reactions in global, modular, and irregular tree networks. The findings clarify the role of network structure in generation of complex transients that can, for example, play a role in intermittent desynchronization of the circadian clock due to external cues or in deep brain stimulations where long transients are required after a desynchronization stimulus.

ABSTRACT. We show how solitary states in a system of globally coupled FitzHugh-Nagumo oscillators can lead to the emergence of chimera states. By a numerical bifurcation analysis of a suitable reduced system in the thermodynamic limit we demonstrate how solitary states, after emerging from the synchronous state, become chaotic in a period-doubling cascade. Subsequently, states with a single chaotic oscillator give rise to states with an increasing number of incoherent chaotic oscillators. In large systems, these chimera states show extensive chaos. We demonstrate the coexistence of many of such chaotic attractors with different Lyapunov dimensions, due to different numbers of incoherent oscillators.

ABSTRACT. Real-world networks are dynamic in that both the state of the network components and the structure of the network (topology) change over time. Most studies regarding network evolution consider either one or the other of these types of network processes. Here we consider the interplay of the two, specifically, we consider how changes in network structure effect the dynamics of the network components. To model the growth of a network we use the specialization model known to produce many of the well-known features observed in real-world networks. We show that specialization results in a nontrivial equitable partition of the network where the elements of the partition form clusters that have synchronous dynamics. The clusters themselves cut across communities and can be thought of as serving the same role within different communities from both a structural and dynamics point of view. We show that these synchronizing clusters inherit their ability to either locally or globally synchronize from the subnetwork from which they are specialized. To the best of the authors' knowledge this may be the first example of a topological mechanism that induces spontaneous synchronization and real-world like growth. Thus, network specialization can be used to model the co-evolution of dynamic and topological features found in real-world systems. (This work is published in Physica A, April 2022.)

ABSTRACT. The optimization problem aiming at the identification of minimal sets of nodes able to drive the dynamics of Boolean networks toward desired long-term behaviors is central for some applications, as for example the detection of key therapeutic targets to control pathways in models of biological signaling and regulatory networks. Unfortunately, the complexity of the optimization problem is exponential, making it exactly solvable on very small systems only. Some scalable approaches exist but they rely on linear approximations; other approaches estimate nonlinear effects but they are generally not scalable. In this talk, we introduce an alternative method inspired by those used in the solution of the well-studied problem of influence maximization for spreading processes in social networks. The computational time of the proposed method scales cubically with the network size. This is achieved thanks to some strong approximations, as for example neglecting dynamical correlations among Boolean variables. However, the method has the desirable feature of fully accounting for the nonlinear nature of Boolean dynamics. We validate the method on small gene regulatory networks whose dynamical landscapes are known by means of brute-force analysis. We then systematically apply it to a large collection of gene regulatory networks revealing that for about 65% of the analyzed networks, the minimal driver sets contain less than 20% of their nodes.

ABSTRACT. The discrete Bass model on two-dimensional Cartesian networks describes the diffusion of new products that spread primarily by spatial peer effects, such as residential photovoltaic solar systems. We show analytically that nodes (residential units) that are located near the town boundary are less likely to adopt than centrally-located ones. This boundary effect is local, and decays exponentially fast with the distance from the boundary. At the aggregate level, the reduction of the adoption level by the boundary scales as~$1/\sqrt{N}$, where~$N$ is the number of nodes. Our theoretical analysis is supported by empirical study of the effect of boundaries on the adoption of solar, in the state of Connecticut. It also provides insight on optimal seeding strategies for promoting solar.

ABSTRACT. In many social, physical, and biological networks, their structure evolves over time with daily, weekly and/or annual cycles. For example, a college's class schedule is usually organized in weekly periodic cycles. Thus motivated, we formulate and analyze a susceptible-infectious-susceptible (SIS) epidemic model over temporal networks with periodically switching connections. Using Floquet theory ---a framework that extends the theory of linear systems to the setting of time-varying periodic systems---we characterize the epidemic threshold and growth/decay rates in terms of a Floquet exponent of a system's monodromy matrix. We further employ this framework to identify and study a Parrondo's paradox for epidemic spreading, whereby a temporal network can have subcritical (epidemic decay) dynamics even if it seemingly appears to be super-critical (epidemic growth) at all instances in time (i.e., ignoring that the network is periodically switching).

ABSTRACT. Does a new product spread faster among heterogeneous or homogeneous consumers? We analyze this question using the stochastic discrete Bass model, in which consumers may differ in their individual external influence rates $\{p_j \}$ and in their individual internal influence rates $\{ q_j \}$. When the network is complete and the heterogeneity is only manifested in $\{p_j \}$ or only in $\{ q_j \}$, it always slows down the diffusion, compared to the corresponding homogeneous network. When, however, consumers are heterogeneous in both~$\{p_j\}$ and~$\{q_{j}\}$, heterogeneity slows down the diffusion in some cases, but accelerates it in others. Moreover, the dominance between the heterogeneous and homogeneous adoption levels is global in time in some cases, but changes with time in others. Perhaps surprisingly, global dominance between two networks is not always preserved under ``additive transformations'', such as adding an identical node to both networks. When the network is not complete, the effect of heterogeneity depends also on its spatial distribution within the network.

ABSTRACT. Connectivity and reachability on temporal networks, which can describe the spreading of a disease, the dissemination of information or the accessibility of a public transport system over time, have been among the main contemporary areas of study in complex systems for the last decade. However, while isotropic percolation theory successfully describes connectivity in static networks, a similar description has not yet been developed for temporal networks. Here, we address this problem and formalize a mapping of the concept of temporal network reachability to percolation theory. We show that the limited-waiting-time reachability, a generic notion of constrained connectivity in temporal networks, displays a directed percolation phase transition in connectivity. Consequently, the critical percolation properties of spreading processes on temporal networks can be estimated by a set of known exponents characterising the directed percolation universality class. This result is robust across a diverse set of temporal network models with different temporal and topological heterogeneities, while by using our methodology we uncover similar reachability phase transitions in real temporal networks too. These findings open up an avenue to apply theory, concepts and methodology from the well-developed directed percolation literature to temporal networks.

ABSTRACT. Over the years, multiple models have been developed to study the spread of smoking. However, these are mainly Ordinary Differential Equation (ODE) models that do not consider all possible interactions between individuals that have been observed empirically. Additionally, they do not take into account the underlying network structure. In this paper, we have developed a network-based Agent-based model that considers all possible interactions. We find that the underlying network structure affects smoking dynamics considerably. In fact, we see that networks with communities embedded in them best replicate the dynamics observed in the real world. Taken together, our model presents a novel approach for developing network-based tobacco control policies.

ABSTRACT. Layered networks have wide range of applications. I show analytically and illustrate numerically, how fluctuations injected in one layer might become amplified in its dependent layers. These results can help building more resilient interdependent networks and avoid major failures.

ABSTRACT. The paper investigates computationally the following research hypotheses: 1. Higher flexibility and discretion in organizational culture results in better mistake management and thus better organizational learning 2. Effective organizational learning requires a transformational leader to have both high social and formal status and consistency. 3. Company culture and leader’s behavior must align for the best learning effects. Computational simulations of the adaptive network were analyzed in different contexts varying in organization culture and leader characteristics. Statistical analysis results proved to be significant and supported the research hypotheses. Ultimately, this paper provides insight into how organizations that foster a mistake-tolerant attitude in alignment with the leader, can result in significantly better organizational learning on a team and individual level.

ABSTRACT. The temporal dynamics of social interactions were shown to influence the spread of disease. Here, we model the conditions of progression and competition for several viral strains, exploring various levels of cross-immunity over temporal networks. We use our interaction-driven contagion model and characterize, using it, several viral variants. We show that the temporal random network and the real-world network provide similar results for the competition under full cross-immunity when considering maxi- mal infection probabilities. When there is no cross-immunity, the dynamics of all three variants are similar over the real-world network, with the faster variants, i.e., with a higher probability of infection, infecting a larger part of the population than the slower ones. The competition dynamics differ substantially when the duration of the meetings is taken as a factor in the contagious process. When variants differ by the minimal time it takes them to infect, with faster variants taking less time to infect than slower variants, the competing conditions become more complex. The slowest variant dies out, leaving the two faster ones to compete. The fast variant dominates a larger part of the network and for a longer duration, followed by the second fastest variant, which creates a large second wave in which it infects at a faster rate than the rate at which that variant infected during the first wave. Thus, we show here that when the duration of meetings is considered in the modeling of temporal networks, the spreading process of competing variants is substantially different.

ABSTRACT. The dynamics of complex systems is most often accessible through the multivariate time series and they are then used to determine the correlaton matrices which, for transparency, are converted into networks, typically reduced to the minimum spanning tree (MST) representation. Of course, by construction the traditional correlation coefficients sizeably compress and reduce the amount of information contained in the original series. The related reduction may result, at the first place, from the fact that such coefficients involve averaging correlations over the whole span of fluctuations and thus do not filter out some possible variability of the intensity of correlations at different amplitudes of fluctuations. Within such an approach, in the case of strong correlations, the resulting MST may give rise to some false signals like for instance promotion a pripheral node to play the role of a central hub. Here the generalization of the concept of cross-correlation coefficient to the q-dependent detrended cross-correlation coefficient is presented such that when varying the q-parameter it acts selectively to cross-correlations between different fluctuation amplitudes at different time scales s of multivariate data. Following such a generalization the family of q-dependent minimum spanning trees (q-MSTs) is introduced which allows to graphically disentangle the composition and organization of correlations and thus to study their varying network characteristics. The utility of such a procedure in addressing the above indicated issues is ilustrated on a recently vital subject of the world cryptocurrency market and of the underlying cross-correlations.

ABSTRACT. Bitcoin is the first and highest valued cryptocurrency that stores transactions in a publicly distributed ledger called the blockchain. Understanding the activity and behavior of Bitcoin actors is a crucial research topic as they are pseudonymous in the transaction network. In this article, we propose a method based on taint analysis to extract taint flows—dynamic networks representing the sequence of Bitcoins transferred from an initial source to other actors until dissolution. Then, we apply graph embedding methods to characterize taint flows. We evaluate our embedding method with taint flows from top mining pools and show that it can classify mining pools with high accuracy. We also found that taint flows from the same period show high similarity. Our work proves that tracing the money flows can be a promising approach to classifying source actors and characterizing different money flow patterns.

ABSTRACT. 1 Introduction Crypto assets are getting a lot of attentions from investors, regulators and policymakers in recent times because of their large market capitalization and volatile nature in price. XRP is one of them developed by Ripple Labs. Inc.. This crypto asset has experienced severe price fluctuations around the end of 2017. The presence of bubbles i.e. explosive price behaviour in this asset has attracted attention from the researchers. Here, we develop a new method of correlation tensor spectra from networks of direct XRP transaction to detect bubbles in XRP price.

2 Data & Method We have collected all the direct transactions between different XRP wallets from October 2, 2017 to March 4, 2018, which is recorded as as ledger data using Ripple Transaction Protocol. We grouped these data into T = 22 weekly XRP networks where wallets are the nodes and a direct transaction from a source wallet to a destination wallet form a directed link between them. The total transaction volume between the wallets represents link-weight. See [1, 2] for the structural properties of the XRP transaction network. We embedded these weighted directed weekly networks using well-known node2vec method [3] with parameters p = q = 1. In these weekly networks, we found that N = 71 nodes are active, i.e. each of these nodes does at least one transaction in every week. We call these 71 nodes as regular nodes. These regular nodes of weekly networks in the embedding space is represented by a d-dimensional vector time series$V_{i}^\alpha (t)$, where $i=1,2,....N$, $t=1,2,3,....T$ and $\alpha=1,2,3,....d$. We have chosen $d = 32$ for our study. Other values of $d$ give qualitatively similar results. The correlation tensor between different components of the regular nodes is defined as $C(t) = \langle V_{i}^\alpha (t) V_{j}^{\beta} (t) \rangle$, where $i,j =1,2,3, ...N$; $\alpha, \beta =1,2,3,...d$; and the $\langle \cdot \rangle$ represents time average for a moving time window of 3 weekly $\{t-1, t, t+1\}$ networks. To find the spectrum of the correlation tensor \begin{equation} C (t) = C_{ij}^{\alpha\beta}(t) = \langle V_{i}^\alpha (t) V_{j}^{\beta} (t) \rangle, \end{equation} we perform a double singular value decomposition (SVD) in the following way:

We carry out the diagonalization of $C_{ij}^{\alpha\beta}$ in terms of $(ij)$-index and $(\alpha\beta)$-index successively by the bi-unitary transformation or equivalently SVD.

The first step is \begin{equation} C_{ij}^{\alpha\beta} = \sum\limits_{k=1}^N L_{ik}\sigma_k^{\alpha\beta} R_{kj}, \end{equation} and the second step is \begin{equation} \sigma_k^{\alpha\beta} = \sum\limits_{\gamma=1}^d \mathcal{L}^{\alpha\gamma} \rho_k^\gamma \mathcal{R}^{\gamma\beta}. \end{equation} Then altogether we have \begin{equation} C_{ij}^{\alpha\beta} = \sum\limits_{k=1}^N \sum\limits_{\gamma=1}^d \rho_k^\gamma (L_{ik} R_{kj}) (\mathcal{L}^{\alpha\gamma} \mathcal{R}^{\gamma\beta}). \end{equation} Here $ \rho_k^\gamma$ is the $N \times d$ generalized eigenvalues. Also note that all eigenvalues are real because the correlation tensor $C$ is symmetric with respect to both node and dimension indices.

3 Results We show the daily XRP/USD exchange rate price from October 2, 2017 to March 4, 2018 in Fig 1 (left). The XRP/USD price has an extraordinary rise and fall between December 2017 and January 2018. This indicates a bubble period for XRP. Our method aims to detect this bubble period from XRP transaction data. For this purpose, we have grouped all the XRP transactions between wallets into 22 weekly networks. The number of nodes in each weekly network is shown in Fig 2 (right). Here also the number of nodes has a sharp rise and fall between December 2017 and January 2018. We have embedded each of these weekly networks using the well-know node2vec algorithm with ordinary random walk method on d = 32 dimensional space. This gives a d-dimensional vector V α i for each node of the networks. From these weekly snapshots of the vectors, we calculate the correlation tensor Cαβ i j (t) for regular nodes as prescribed in the section 2. The correlation tensor has N × d eigenvalues. While most of the eigen- values are very small, there are few extreme eigenvalues which deviates from the bulk of the eigenvalues. The largest eigenvalue λ1 is generally associated with the market mode. We show the variation of λ1 in Fig 2 (left). It shows a sharp fall in its magnitude even before the bubble period and a rise to the pre-bubble period values later. So this largest eigenvalue can be useful as an early warning signal for the crashes in price. We have observed qualitatively similar behaviour for the second largest eigenvalue λ2. The spectral gap as shown in Fig 2 (right) also gives a similar indication of crashes in XRP price. How- ever, we have noticed that the third and fourth largest eigenvalues do not give such an indication. Summary. We have devolved a new method of correlation tensor spectra from XRP transaction networks, which can be used as an early warning signal for crashes in XRP price. Although we have studied it only for XRP, this method is very general and can be applied to other trading assets. References 1. Ikeda, Y.: Characterization of XRP Crypto-Asset Transactions from Networks Scientific Ap- proach. In: Aruka, Y. (eds) Digital Designs for Money, Markets, and Social Dilemmas. Evo- lutionary Economics and Social Complexity Science, Springer, Singapore vol 28, (2022). 2. Aoyama H. Fujiwara, Y., Hidaka, Y., Ikeda, Y.: Cryptoasset Networks: Flows and Regular Players in Bitcoin and XRP. Complex networks 2021, book of abstracts pp. 437-439 (2021). 3. Grover, A., Leskovec, J.: node2vec: Scalable feature learning for networks. In Proceedings of the 22nd ACM SIGKDD international conference on Knowledge discovery and data mining pp. 855-864 (2016).

ABSTRACT. Time series analysis has proven to be a powerful method to characterize different phenomena in biology, neuroscience, economics, and to understand some of their underlying dynamical features. However, to date it remains unclear whether the information encoded in multivariate time series, such as the evolution of financial assets traded in major financial markets, or the neuronal activity in the brain, stems from either independent, pairwise, or group interactions (i.e., higher-order structures [1]).

In this work, we propose a novel framework to characterize the instantaneous co-fluctuation patterns of signals at all orders of interactions (pairs, triangles, etc.), and to investigate the global topology of such co-fluctuations [2]. In particular, after i) z-scoring the N original time series, ii) we calculate the element-wise product of the z-scored time series for all the possible k-order patterns (i.e. edges, triplets, etc). Here, the generic elements represent the instantaneous co-fluctuation magnitude between a (k+1) group interaction. To distinguish concordant group interactions from discordant ones in a k-order product, concordant signs are always positively mapped, while discordant signs are negatively mapped (Fig. 1a-b). iii) The resulting new set of time series encoding the k-order co-fluctuations are then further z-scored across time, to make products comparable across k-orders. For each time frame t, a weighted simplicial complex condenses all the k-order co-fluctuations (Fig. 1c). iv) Finally, we construct a weight filtration, by sorting all the k-order co-fluctuations by their weights. The weight filtration proceeds from the top down − in the spirit of persistent homology − so that when k-order patterns are gradually included, weighted holes and cliques start to appear (i.e., descending from more coherent patterns to less coherent). Yet, to maintain a well-defined weight filtration, only k-order patterns respecting the simplicial closure condition are included, while the remaining ones are considered as simplicial violations, or “hypercoherent” states, and analysed separately (Fig. 1d).

We show that the instantaneous persistent topological properties and the number of violations uncovered by our framework distinguish different regimes of coupled chaotic maps [3]. This includes the transition between different dynamical phases and various types of synchronization (Fig. 2a). Armed with these interpretational benchmarks, we also apply our method to resting-state fMRI signals and to financial time series (Fig. 2b). We find that, during rest, the human brain mainly oscillates between chaotic and partially intermittent states, with higher-order structures mainly reflecting somatosensory areas (Fig. 2c top). In financial time series, by contrast, higher-order structures discriminate crises from periods of financial stability (Fig. 2c bottom).

Overall, our approach suggests that investigating the higher-order structure of multivariate time series might provide new insights compared to standard methods, and might allow us to better characterise group dependencies inherent to real-world data.

ABSTRACT. Decentralized Finance (DeFi) stands for a new paradigm that aims to disrupt established financial services. It offers services in the form of smart contracts, which are executable software programs deployed on top of blockchains such as Ethereum. Despite being a relatively recent development, we can already observe rapid growth in DeFi protocols enabling lending of cryptoassets, exchanging them for other cryptoassets without intermediaries, or betting on future price developments. In this study we analyze the dependency network of the DeFi ecosystem and quantify the systemic risk of the highest degree smart contracts. We find very few high systemic risk smart contracts that can potentially threaten the functioning of almost the entire DeFi ecosystem. For the three major stablecoins we find that USDT and USDC are of significant relevance to the network, but do not belong to highest risk smart contracts. Our method for estimating the systemic risk of smart contracts is highly relevant to regulators as it provides a way to direct policy measures on the riskiest contracts and protocols.

ABSTRACT. We consider a network-based framework for studying causal relationships in financial markets and demonstrate this approach by applying it to the entire U.S. stock market. Directed networks (referred to as ``causal market graphs'') are constructed based on publicly available stock prices time series data during 2001-2020, using Granger causality as a measure of pairwise causal relationships between all stocks. We consider the dynamics of structural properties of the constructed network snapshots, group stocks into network-based clusters, as well as identify the most ``influential'' market sectors via the PageRank algorithm. Interestingly, we observed drastic changes of the considered network characteristics in years that corresponded to significant global-scale events, most notably, the financial crisis of 2008 and the COVID-19 pandemic of 2020.

This presentation is based on the published paper: Shirokikh, Oleg, Pastukhov, Grigory, Semenov, Alexander, Butenko, Sergiy, Veremyev, Alexander, Pasiliao, Eduardo L. and Boginski, Vladimir. "Networks of causal relationships in the U.S. stock market" Dependence Modeling, vol. 10, no. 1, 2022, pp. 177-190. https://doi.org/10.1515/demo-2022-0110

ABSTRACT. Symbolic transfer entropy is a powerful non-parametric tool to detect lead-lag between time series. Because a closed expression of the distribution of Transfer Entropy is not known for finite-size samples, statistical testing is often performed with bootstraps whose slowness prevents the inference of large lead-lag networks between long time series. On the other hand, the asymptotic distribution of Transfer Entropy between two time series is known. In this work, we derive the asymptotic distribution of the test for one time series having a larger Transfer Entropy than another one on a target time series. We then measure the convergence speed of both tests in the small sample size limits via benchmarks. We then introduce Transfer Entropy between time-shifted time series, which allows to measure the timescale at which information transfer is maximal and vanishes. We finally apply these methods to tick-by-tick price changes of several hundreds of stocks, yielding non-trivial statistically validated networks.

ABSTRACT. In this presentation, we would like to show how the method of Non-negative matrix factorization and tensor factorization can be applied to two economic networks; the one of cryptoasset of Bitcoin [1], and the one of a regional bank's accounts in Japan [2].

[1] Fujiwara, Y. and Islam, R.: Bitcoin's Crypto Flow Network. JPS Conference Proceedings, Vol.~36, 011002 (2021) [2] Fujiwara, Y., Inoue, H., Yamaguchi, Y., Aoyama, H., Tanaka, T., Kikuchi, K.: Money flow network among firms' accounts in a regional bank of Japan. EPJ Data Science, 10, 19 (2021)

ABSTRACT. Early warning signals—statistics that can indicate proximity of a system to bifurcation—have been developed for univariate and multivariate data. Multivariate early warning signals tend to treat a networked system as a unified entity with one major transition to predict, unable to reliably detect additional transitions after the first. However, multistage transitions, in which some but not all nodes have changed state at equilibrium for a range of control parameter values, are common in models of dynamics on networks, and testing for multistage transitions in empirical settings may be important for several applications. We develop early warning signals in networked systems that can (a) detect multistage transitions, and (b) do so with reduced numbers of observed nodes.

ABSTRACT. Evolutionary dynamics on networks has been an active research topic. The spread of a mutant in a given population of residents can be substantially affected by the population structure. The mutants generally have a fitness difference with the resident population, which makes them either more or less likely to produce offsprings. The networks that amplify the difference in fitness, thus making it more likely for advantageous mutants to spread and disadvantageous mutants to die out, are called amplifiers of selection; the networks that suppress this difference in fitness are called suppressors of selection. The network on which evolutionary dynamics are occurring may be time-varying. Note that the evolutionary dynamics with fixed fitnesses for the resident and mutant types are biased voter models that mimic stochastic opinion formation on networks; contact networks underlying such biased opinion competition may be dynamic. However, there exists little literature on evolutionary dynamics on temporal networks.We introduce evolutionary dynamics on switching temporal networks in which two fixed networks G1 and G2 periodically alternate. We surprisingly found that there were many suppressors for switching networks on 6 nodes, whereas only one suppressor is exists for static networks on 6 nodes .

ABSTRACT. Time-varying higher-order interactions among more than two units are typically modeled as a temporal higher-order network in the framework of simplicial complex. For the temporal evolution of the higher-order structure of human proximity interactions in five different social settings, previous work found the characteristics of the configuration transitions in a temporal higher-order network before and after triplet interaction events occur. Recently, food science and computing have been attracting attention due to the increasing popularity of recipe sharing services in social media. In this paper, aiming to reveal the characteristics of the temporal evolution of homemade recipes in terms of combinations of ingredients, we propose a method of analyzing the configuration transitions in a temporal higher-order network of ingredients before and after new higher-order links are formed in the framework of temporal simplicial complex. Using real data of a Japanese recipe sharing service, we empirically demonstrate the effectiveness of the proposed method, and apply it to analyzing the dynamical properties of higher-order networks of ingredients for Japanese homemade recipes.

ABSTRACT. Numerous studies over the past decades established that real-world networks typically follow preferential attachment and detachment principles. Subsequently, this implies that degree fluctuations monotonically increase while rising up the “degree ladder”, causing high-degree nodes to be prone for attachment of new edges and for detachment of existing ones. Despite the extensive study of node degrees (absolute popularity), many domains consider node ranks (relative popularity) as of greater importance. This raises intriguing questions - what dynamics are expected to emerge when observing the ranking of network nodes over time? Does the ranking of nodes present similar monotonous patterns to the dynamics of their corresponding degrees? In this study we show that surprisingly the answer is not straightforward. By performing both theoretical and empirical analyses, we demonstrate that preferential principles do not apply to the temporal changes in node ranking, and these rather follows a non-monotonous inverse-U shaped curve. These findings provide plausible explanations to observed yet hitherto unexplained phenomena, such as how superstars fortify their ranks despite massive fluctuations in their degrees, and how stars are more prone to rank instability.

ABSTRACT. Entropy metrics are nonlinear measures to quantify the complexity of time series. Among them, permutation entropy is a common metric due to its robustness and fast computation. Multivariate entropy metrics techniques are needed to analyse data consisting of more than one time series. To this end, we present a multivariate permutation entropy, MPEG, using a graph-based approach.

Given a multivariate signal, the algorithm MPEG involves two main steps: 1) we construct an underlying graph G as the Cartesian product of two graphs G1 and G2, where G1 preserves temporal information of each times series together with G2 that models the relations between different channels, and 2) we consider the multivariate signal as samples defined on the regular graph G and apply the recently introduced permutation entropy for graphs.

Our graph-based approach gives the flexibility to consider diverse types of cross channel relationships and signals, and it overcomes with the limitations of current multivariate permutation entropy.

ABSTRACT. This paper aims to understand to what extent the amount of drug (e.g., cocaine) trafficking per country can be explained and predicted using the global shipping network. We propose three distinct network approaches, based on topological centrality metrics, Susceptible-Infected-Susceptible spreading process and a flow optimization model of drug trafficking on the shipping network, respectively. These approaches derive centrality metrics, infection probability, and inflow of drug traffic per country respectively, to estimate the amount of drug trafficking. We use the amount of drug seizure as an approximation of the amount of drug trafficking per country to evaluate our methods. Specifically, we investigate to what extent different methods could predict the ranking of countries in drug seizure (amount). Furthermore, these three approaches are integrated by a linear regression method in which we combine the nodal properties derived by each method to build a comprehensive model for the cocaine seizure data. Our analysis finds that the unweighted eigenvector centrality metric combined with the inflow derived by the flow optimization method best identifies the countries with a large amount of drug seizure (e.g., rank correlation .45 with the drug seizure). Extending this regression model with two extra features, the distance of a country from the source of cocaine production and a country's income group, increases further the prediction quality (e.g., rank correlation .79). This final model provides insights into network derived properties and complementary country features that are explanatory for the amount of cocaine seized. The model can also be used to identify countries that have no drug seizure data but are possibly susceptible to cocaine trafficking.

ABSTRACT. Collecting Spatio-temporal data via wearable sensors enables unobtrusive studies on individuals' behavior in various applications. Recently, interest has been growing in using wearable data to analyse the social network of individuals such as children's social network in a schoolyard. In fact, social network analysis identifies an individuals' position and interactions in a network. The static network metrics, however, cannot capture the complex dynamics of such data. To study and understand the social phenomenon, one needs to carefully devise new metrics that consider both time and space. While temporal metrics have been proposed, spatial metrics on such temporal dynamics have not been considered. This article aims to introduce a new metric on temporal dynamics of social networks in which both temporal and spatial aspects of the network are considered for measuring the social dynamics of the individuals. The main strength of this metric lies in examining the environmental characteristics when analyzing individual's social behavior. We evaluated our metric in the real-world by collecting a Spatio-temporal dataset from a primary special education school. We find that our temporal metric uniquely quantifies children’s relationships over time concerning the spatial properties of the schoolyard.

ABSTRACT. Integrating complex network theory, link prediction theory, and related research on industrial competition relationship, this paper proposes the theoretically analytical framework of the competitive relationship among Z-Park high-tech enterprises. By constructing a link prediction model, we reveal the internal dynamics that affect the evolution of the competitive network of enterprises, seek the best index reflecting the network formation mechanism, and apply it to the prediction of potential competitive associations.

ABSTRACT. This talk explores network and crowds' physical, behavioral, and computational limits for solving time-critical problems. I describe several real-world episodes in which we used social networks to mobilize the masses to deter threats of unprecedented complexity. From finding red weather balloons hidden across the United States, to tracking down thieves in a global hunt, to reconstructing shredded classified documents, to disabling a harmful Artificial Intelligence, the potential of crowdsourcing is real, but so are the exploitation, sabotage, and polarization dynamics that undermine the power of crowds. Keeping our communities, organizations, and institutions safe depends on harnessing the networks and crowds' power and talent while defending themselves from their aggressors.

ABSTRACT. In the last decades, network theory has been successfully applied to explore a great variety of complex systems. Particularly interesting is the study of network local patterns that can shed light on the emergence of global properties. In this perspective, network binary motifs represent a fruitful example. Motifs are set of nodes and edges, completely described by their size and type of links (directed/undirected). It is well known that the introduction of edge-values can offer novel insights about the system properties; however, only few works have proposed an extension of the motif concept to the weighted case. The method proposed in this work has few ingredients: (i) unbalanced weighted networks; (ii) a sink node and (iii) a random walker with a limited number of steps. The sink node is introduced to compensate the excess of ingoing flows and balance the network. It allows to highlight the role of weights heterogeneity in shaping the network structural organization in subpatterns. The approach gives the frequency of paths of any possible length observable within a fixed number of steps of a random walker placed on an arbitrary node. In other terms, all subgraphs of dimension smaller than and equal to the maximum number of steps can be considered mutually exclusive events, i.e., their total occupation probability sums up to one. We applied our approach to different real networks and test the significance of weighted motifs occurrences using a randomization technique based on the entropy maximization of the system under local constraints: degree and strength (Enhanced Directed Configuration Model, EDCM). We consider a maximum of 3 steps offering both analytical results and simultations outcome. In principle, we can include any number of steps, but this exponentially increases the computational complexity of the problem.

ABSTRACT. We consider entities interacting over time: individuals meeting, customers buying products, etc., each entity being labeled with some information that may depend on time, and possibly extracted from the interaction nature. Capturing the dynamics as well as the structure of these interactions is of crucial importance for analysis. We are interested here in mining sequences of such interactions. For that purpose, we define core closed patterns in this context and introduce algorithms to enumerate them on a labeled stream graph. We run experiments on two real-world datasets, one representing interactions among students and the other representing citations between authors.

ABSTRACT. Interactions among objects are usually modelled using graphs. Nevertheless, these relations may change over time and there exist different kind of relations among object that need to be integrated. We introduce a new network model, called temporal dual networks, to deal with interactions that changes over time and to integrate information coming from two different networks. We consider a fundamental problem in graph mining, that is finding densest subgraphs on this new model. We propose an approach based on both network alignment and dynamic programming. Given two temporal graphs, we obtain a dual temporal graph via alignment and then we look for densest subgraphs in the obtained graph. We present a dynamic programming algorithm to solve the problem in polynomial time. Since this algorithm is not applicable even to medium size network, we present a heuristic that is based on (1) constraining the dynamic programming to consider only bounded temporal graphs and (2) a local search procedure. We show that our method is able to return optimal or near optimal solution even for temporal graphs having 10000 vertices and 10000 timestamps.

ABSTRACT. Many complex systems exist in the physical world and therefore can be modeled by networks in which their nodes and edges are embedded in space. However, classical network motifs only use purely topological information and disregard other features. In this paper we introduce a novel and general definition subgraph abstraction that incorporates spatial information, therefore enriching their characterization power. Moreover, we describe and implement a method to compute and count our spatial subgraphs in any given network. We also provide initial experimental results by using our methodology to produce spatial fingerprints of real road networks, showcasing its discrimination power and how it captures more than just simple topology.

ABSTRACT. Association Football is probably the world’s most popular sport. Being able to characterize and compare football players is therefore a very important and impactful task. In this work we introduce spatial flow motifs as an extension of previous work on this problem, by incorporating both temporal and spatial information into the network analysis of football data. Our approach considers passing sequences and the role of the player in those sequences, complemented with the physical position of the field where the passes occurred. We provide experimental results of our proposed methodology on a real-life event data from the Italian League, showing we can more accurately identify players when compared to using purely topological data.

ABSTRACT. The Mean Hitting Time is a fundamental structural measure of random walks on networks with many applications ranging from epidemic diffusion on networks to fluctuations in stock prices. It measures the mean expected time for a random walker to reach all the source-destination pair nodes in the network. Previous research shows that it scales linearly with the network size for small-world sparse networks. Here, we calculate the Mean Hitting Time for large real-work complex networks and investigate how it scales with the $q$-subdivision operation used to grow the network. Indeed, this operation is essential in modeling realistic networks with small-world, scale-free and fractal characteristics. We use the Eigenvalues and eigenvectors of the normalized adjacency matrix of the initial network $G$ to calculate the Hitting Time $T_{ij}$ between nodes $i$ and $j$. We consider two complex real-world networks to analyze the evolution of the Mean Hitting Time as the networks grow with the $q$-subdivision. Results show that the Mean Hitting Time increases linearly with the value of $q$. This work provides insight into the design of realistic networks with short Mean Hitting Time.

ABSTRACT. Alzheimer’s disease (AD) is a brain disconnection syndrome, where functional connectivity analysis can detect changes in neural activity in pre-dementia stages. Functional connectivity networks – from functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) – are susceptible to signal noise from biologic artefacts (e.g. cardiac artefacts) and environmental sources (e.g. electrical interference). A particular challenge for EEG is volume conduction, whereby a signal from a single source propagates through biological tissue to be detected simultaneously by multiple sensors (channels). The imaginary part of coherency (iCOH) provides a measure for connectivity that avoids this signal contamination, by ignoring correlation between signals with zero or π-phase lag. This removes false instantaneous activity but comes at the cost of erasing true instantaneous activity as well. We propose a network assessment, eigenvector alignment (EA), that is robust to noise and connectivity erasure by evaluating pairwise relationships (alignments) from the pattern of functional connectivities. These patterns are captured through eigenvector embedding of the connectivity network nodes in a Euclidean space. This analysis of pattern rather than connectivity magnitudes, allows subjects to be directly compared from different test conditions and suggests that EA can provide a robust basis for detecting neural changes that are key in understanding and treating a wide variety of conditions

ABSTRACT. Falsification is the basis for testing existing hypotheses, and a great danger is posed when results incorrectly reject our prior notions (false positives). Though nonparametric and nonlinear exploratory methods of uncovering coupling provide a flexible framework to study network configurations and discover causal graphs, multiple compar- isons analyses make false positives more likely, exacerbating the need for their control. We aim to robustify the Gaussian Processes Convergent Cross-Mapping (GP-CCM) method through Variational Bayesian Gaussian Process modeling (VGP-CCM). We alleviate computational costs of integrating with conditional hyperparameter distributions through mean field approximations. This approximation model, in conjunction with permutation sampling of the null distribution, permits significance statistics that are more robust than permutation sampling with point hyperparameters. Simulated unidirectional Lorenz-Rossler systems as well as mechanistic models of neurovascular systems are used to evaluate the method. The results demonstrate that the proposed method yields improved specificity, showing promise to combat false positives.

ABSTRACT. Artificial Neural Networks (ANN), and in particular Deep Learning (DL), are biologi- cally inspired computing systems that represents the state-of-the-art AI algorithms. In the last years several evidences support the notion that non-random topologies can lead to an improvement of the performance of ANNs (see Carroll, T. L. and Pecora, M. L., Chaos:An Interdisciplinary Journal of Nonlinear Science, (2019) and Damicelli, F., Hilgetag, C., and Goulas, A., bioRxiv, (2021) ). In our analysis, 45 connectomes sampled from brain regions of several organisms are used to build the reservoir. The performances of such connectome-based RC models are tested on a one-step-ahead prediction task with different input signals. Random signals and deterministic chaotic signals are considered. Furthermore, for each connectome analogous synthetic networks are generated as null models of increasing complexity to verify if the behaviour of the empirical networks actually depends on topological correlations. Erdo ̋s Re ́nyi (ER) model, Configuration Model (CM) and Stochastic Block Model (SBM) are used to generate synthetic networks. We find a correlation between RC performances and both number of nodes and rank of the covariance of a matrix. Moreover, our analysis reveals the ability of the neural system to perform significantly better according to specific network topological measures. In particular, it emerges that the modularity and the average degree of the network play a key role in model accuracy; higher values of modularity and average degree can significantly affect RC performances. Importantly, RC performances are significantly improved or reduced depending on the stochastic or deterministic nature of the signals.

ABSTRACT. Functional connectivity (FC) network} analysis has become one of the leaking frameworks for the study of the brain as a complex system. Despite its widespread adoption, there is an inherent limitation that is rarely discussed: the only dependencies visible to the pairwise correlation measure are bivariate ones: there are no direct ways to infer statistical dependencies between three or more variables. Recent work in multivariate information theory has lead to the development of a variety of tools for recognizing statistical dependencies beyond pairwise correlations

We show partial entropy decomposition, a generalization of the more widely known partial information decomposition, reveals two fundamental limitations of the widely-used functional connectivity network model. In human fMRI data, we found a large amount of higher-order information and synergistic information. Redundant information compromises the specificity of any pairwise edge, while synergistic information shows that bivariate network analyses are incomplete. Standard FC features such as edge density, clustering coefficient, and community structure are biased in favor of redundancy and against synergy, making it difficult to distinguish between disintegrated systems, and integrated, but highly synergistic ones. These results suggest that there is a large space of higher-order interactions in the brain to explore and which may reveal many heretofore unseen links between brain structure, dynamics, and function.

ABSTRACT. Understanding how the network structure of cortical microcircuit affects the computations that the brain can perform is one of the key challenges in neuroscience. In this work we explore the network architecture of the cortical microcircuit in the human medial temporal lobe by patch-clamping neural tissue from epilepsy patients.We then perform a network motif analysis and compare our networks with what is known about somatosensory cortex in rodents. We demonstrate that the structural features that differentiate the middle temporal gyrus from humans and the somatosensory cortex from rodents can increase the information capacity of the network dynamics. Furthermore, we show that this can help improve the performance of artificial recurrent neural networks in behaviorally relevant tasks.

ABSTRACT. Network models of the brain are among the most powerful tools for understanding its structure and function. Recently, neuroscientists and complexity scientists have begun studying higher-order interactions between multiple brain regions. In this paper, we use multivariate information theory to show that the brain has a large number of higher-order, synergistic subsystems that are invisible when considering a pairwise graph structure. We find subsystems that are maximally synergistic and relate these subsystems to pairwise FC values and canonical functional networks. Our analysis reveals a vast, unexplored higher-order structure in the brain.

ABSTRACT. Models of human navigation have been investigated in many ways on complex networks. These findings suggest that the characteristics of human navigation change during the navigation from the start to the destination. However, it is not fully clear to what extent the navigation is defined by the human navigator or the graph and the environment. Our work examines the early phase of human navigation, where we investigate the impact of the graph structure on human navigation with a random walk model based on PageRank. Our results suggest that a very high portion of human navigation in the early generalization phase can be modeled with random navigation.

ABSTRACT. University rankings are increasingly adopted for academic comparison and success quantification, even to establish performance-based criteria for funding assignment. However, rankings are not neutral tools, and their use frequently overlooks disparities in the starting conditions of institutions. In the present work, we establish a method to detect a bias generically affecting the scores of academic institutions in rankings, quantify its effect on each single university, and decouple it to obtain a fairer rating system. The key idea consists in partitioning universities in similarity groups, determined from multifaceted data using complex network analysis, and referring the performance of each institution to an expectation based on its peers. We have observed that the territorial context is relevant in determining the performance in rankings, prompting an effect that can trigger self-reinforcing awarding mechanisms. Debiasing thus provides interesting insights to design fairer strategies for performance-based funding allocation. Future work will be devoted to the complementary point of view, namely the analysis of the spillover effect of outstanding universities on the territory in which they are embedded.

ABSTRACT. In recent years, well-being and mental health concerns for PhD students have been increasing. According to a recent survey conducted in 2019 by Nature on 6,300 PhD students, 36% responded that they sought help for anxiety or depression caused by their studies (Woolston, 2019). Another devastating fact is that doctoral students are 2.43 times more likely to have a common psychiatric disorder than the rest of the highly educated population (Levecque et al. 2017). It is therefore important to examine the journey of doctoral students not only through the lens of academic "success measures" such as publication numbers, citation counts, fellowships received etc. but also at their overall well-being and the quality of the environment that supports them in fulfilling their potential.

Acknowledgements contain such profound details of their authors' academic journey and environment; however, research efforts to study how they vary concerning disciplinary and demographic differences have remained limited. To fill this gap, we investigated 26,236 PhD dissertations, obtained from ProQuest Open Access Dissertations & Theses database, 99% of which are from the United States in the last 20 years. We systematically identified acknowledged individuals and institutions, by using a data-driven approach supported by manual inspection and found the entities in the acknowledgements such as mother, brother, advisor, colleague etc. We then enriched our dataset by adding the number of publications before and after the graduation of associated PhD students from Dimensions.ai API to be able to examine the relationship between the characteristics of acknowledgements and productivity levels of PhD students.

Firstly, we uncovered the academic support network. For this task, we employed a shallow but efficient neural network model, Doc2Vec, to represent each support provider in a vector format by looking at the verbs, adverbs, nouns and adjectives that are used to acknowledge them. Then, we used these vectors to create an "academic support network" by applying the Girvan-Newman community detection algorithm. It revealed five distinct communities that support students along the way: Academic, Administration, Family, Friends \& Colleagues, and Spiritual.

Secondly, we compared gender and disciplines in terms of who and how they are acknowledging. We showed that female students, compared to their male counterparts, mention fewer people from each of these communities except for their families. However, sentiment analysis results demonstrated that they use a more positive language when acknowledging each community. Our results also suggested that the total number of people mentioned in the acknowledgements allows disciplines to be represented on an individual science - team science plane as their magnitudes change, in which case the Social Sciences & Humanities and Life Sciences & Humanities are on the associated extremities, respectively.

Lastly, we applied an Inverse Gaussian regression model to describe the relationship between students' demographics, their acknowledgement characteristics and productivity levels. Results indicated that women's productivity level is slightly lower than that of males when considering the number of publications alone. This is critically important because it means that studying the doctoral process may help us better understand the adverse conditions women face early in their academic careers. We also showed that male students who mention more people from their academic community are associated with higher levels of productivity. University rankings are found to be positively correlated with productivity and the size of academic support networks. However, neither university rankings nor students' productivity levels were correlated with the sentiments students express in their acknowledgements. Our results point to the importance of academic support networks by explaining how they differ and how they influence productivity.

ABSTRACT. The principle of the adjacent possible describes how new concepts can be discovered or invented when they are 'one step away' from already discovered concepts. However, what is meant by 'one step away' is not always clear. In this contribution, we formulate the principle of the adjacent possible in terms of a network model of conceptual discovery, thereby allowing for a precise definition of 'one step away'. Apart from reproducing stylized facts such as Heap's law and Zipf's law, our model provides insight into the relation between knowledge structure and knowledge generation, and how this relation may vary across different fields of science and technology.

ABSTRACT. In this work, we analyze the polarization configuration of voting behavior in several parliamentary systems across political policy domains. We specifically look at the parliamentary systems of the US, Germany, Finland, and Ireland over a range of election periods. We propose two approaches to study this system. In the "mesoscopic view" in which the polarization of the system and its properties is investigated by quantifying the relationship between the behavior of the political parties. In the "microscopic view" on the other hand, we build networks of co-voting behavior of the members of the parliament (MPs), and using community detection and multi-layer network analysis techniques we look into polarization behavior. Specifically, we investigate if the polarization of parties is "aligned" across different domains of policy making.

In both approaches, we observe periods when some antagonistic parties kept the same polarization intensity across many different domains. We also discover periods of highly aligned and highly polarized political systems which could have/might be harming responsible and constructive policy-making and social discourse. Our initial results demonstrate consistency between mesoscopic and mesoscopic analyses in most cases and also highlight the instances these views differ.

ABSTRACT. Hierarchical relationships are pivotal for social structures in human beings. Yet, little is known about mechanisms connecting social hierarchies and friendship in elementary school children. Here, we present the results of a large-scale experiment (856 children between 9 to 12 years) from 14 different elementary schools in Santiago de Chile, designed to measure social status through aggregated cooperative patterns and then explore the connection between friendship and dyadic cooperation in students with different social status. We map each classroom's cooperative network using a modified Prisoner's Dilemma in a lab in the field setting. In a networked set up of tablet computers, each student played the game in pairs with each classmate. They had to decide simultaneously how many tokens (between o to 10) send to their peers in each round. Thus, we proxy social status using the page rank, which considers the total number of received tokens and the sender's social position. To measure friendship, we run a peer nomination questionnaire, where students nominated up to five friends. Then, we investigate how dyadic cooperation varies according to social status and friendship. In general, we find that the more significant the difference in social status, the greater the dyadic cooperation gap, indicating acts of deference from lower-status individuals to higher-status individuals. However, when we separately analyze relationships involving mutual declarations of friendship, the association between social status and cooperation disappears. Among friends, we do not observe acts of deference from the lower status to the higher status member of the dyad. These results suggest that friendship implies fundamental equality, which is not affected by social status differences in elementary school students.