Abstract : In recent years, research on graphs and networks has received much attention, and it has emerged in the application of channel coding, biomedicine, social governance, and other fields. The mini-symposium will focus on structural analysis and dynamics modelling in complex networks, including influence maximization of high-order networks, evolutionary games, etc.
[03989] Evolutionary Game Theory on Dynamic Networks
Format : Online Talk on Zoom
Author(s) :
Qi Su (University of St Andrews)
Alex McAvoy (University of Pennsylvania)
Joshua B. Plotkin (University of Pennsylvania)
Abstract : The study of the evolution of cooperative behavior on static networks helps to understand how population structure can facilitate the spread of prosocial traits. However, real-world interactions are usually transient and subject to external restructuring, making the study of strategic behavior on dynamic networks difficult. This study provides an analytical treatment of cooperation on dynamic networks and demonstrates that transitions among network structures can promote the spread of cooperation, even if individual social networks inhibit it when static. The findings highlight the significant impact of dynamic social structures on the evolution of prosocial traits.
[02059] An efficient adaptive degree-based heuristic algorithm for influence maximization in hypergraphs
Format : Online Talk on Zoom
Author(s) :
Xiu-Xiu Zhan (Hangzhou Normal University)
Abstract : Influence maximization (IM) has shown wide applicability in immense fields over the past decades. Previous researches on IM mainly focused on the dyadic relationship but lacked the consideration of higher-order relationship between entities, which has been constantly revealed in many real systems. An adaptive degree-based heuristic algorithm, i.e., Hyper Adaptive Degree Pruning (HADP) which aims to iteratively select nodes with low influence overlap as seeds, is proposed in this work to tackle the IM problem in hypergraphs. Furthermore, we extend algorithms from ordinary networks as baselines. Results on 8 empirical hypergraphs show that HADP surpasses the baselines in terms of both effectiveness and efficiency with a maximally 46.02% improvement. Moreover, we test the effectiveness of our algorithm on synthetic hypergraphs generated by different degree heterogeneity. It shows that the improvement of our algorithm effectiveness increases from 2.66% to 14.67% with the increase of degree heterogeneity, which indicates that HADP shows high performance especially in hypergraphs with high heterogeneity, which is ubiquitous in real-world systems.
[04046] Identifying vital nodes through augmented random walks on higher-order networks
Format : Online Talk on Zoom
Author(s) :
Xiao-Long Ren (Yangtze Delta Region Institute (Huzhou), University of Electronic Science and Technology of China)
Abstract : Identifying vital nodes is still unclear in the study of higher-order networks. We introduce a multi-order graph obtained by incorporating the higher-order bipartite graph and the classical pairwise and propose a Higher-order Augmented Random Walk model. Our model provides a new approach for ranking nodes at multiple scales. Our method outperforms other indicators in identifying vital nodes and can scale to various tasks in complex networks, including the information spread maximization and network dismantling problem.
[04449] Emergence of Cooperation Through Coevolving Time Scale in Spatial Prisoner's Dilemma
Format : Online Talk on Zoom
Author(s) :
Zhihai Rong (Donghua University)
Abstract : Understanding the emergence of cooperation is a challenging problem and has drawn a wide attention from various fields including sociology, economics and biology. Evolutional game theory provides a powerful framework for investigating this problem. In the traditional networked evolutionary game theory, most researchers usually assume an individual will immediately update its strategy after one round of game with its neighbors. However, in the social and biological systems the strategy-selection time scale may be slower than the interaction time scale, i.e., an individual will hold its current strategy and play several rounds of game with its neighbors, and then update its behavior. In this talk, I will introduce some results about coevolving time scale in spatial Prisoner’s dilemma. When the individuals can adjust their strategy-selection time scale according to some rules, optimal cooperation can be induced by proper adaptive rate in the strategy-selection time scale. The results are analyzed through the spatial pattern and feedback mechanism of individual behavior. This investigation may have potential implications in the design of consensus protocol in multi-agent systems.
[01988] information-opinion dynamics on social multilayer networks
Format : Online Talk on Zoom
Author(s) :
Fei Jing (City University of Hong Kong)
Abstract : Here we model these two phenomena as a co-evolution dynamics of information and public opinion on heterogeneous multiplex networks, including a few extreme individuals with constant opinions and a vast majority of general individuals with vacillating views.
[04619] Characterizing Cycle Structure in Complex Networks
Format : Online Talk on Zoom
Author(s) :
Tianlong Fan (University of Fribourg)
Abstract : A cycle is the simplest structure that brings redundant paths in network connectivity and feedback effects in network dynamics. In this work, we define the cycle number matrix, and the cycle ratio, an index that quantifies node importance. Numerical experiments on identifying vital nodes for network connectivity and synchronization and maximizing the early reach of spreading show that the cycle ratio performs overall better than other benchmarks.
[03130] Collaborative deep learning framework for network inference and dynamical prediction
Format : Online Talk on Zoom
Author(s) :
Xiao Ding (Anhui University)
Abstract : How to use incomplete data to infer network structure as well as predict the dynamics simultaneously is a meaningful and challenging question. To this end, we develop a COllaborative deep learning framework for Network inference and Dynamical prediction (CoND). Extensive experiments demonstrate that COND outperforms the baseline methods regarding both tasks for different networks and dynamical models. To further validate the effectiveness of COND, we demonstrate the superior performance of CoND on two real datasets.
[05541] Multichannel game on structured populations
Format : Online Talk on Zoom
Author(s) :
Fanpeng Song (Shandong University)
Abstract : In this talk, we introduce a framework of multichannel game on structured populations and explore the effect of topological properties to the evolutionary dynamics. Significantly, we find that the heterogeneity of populations is detrimental to the dynamics, especially in BA networks. In addition, modest population size and high interaction are in favor of the dynamics. These results are meaningful for the research on cooperation and human development.
