Modeling Dynamic Network with Centrality-based Logistic Regression

Dynamic Networks;Network Analysis;Centrality;Exponential Random Graphs;Statistics

Kulmatitskiy, Nikolay

University of Waterloo

关键词: Dynamic Networks;    Network Analysis;    Centrality;    Exponential Random Graphs;    Statistics;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/6290/1/Kulmatitskiy_Nikolay.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

Statistical analysis of network data is an active field of study, in which researchers inves-tigate graph-theoretic concepts and various probability models that explain the behaviourof real networks.This thesis attempts to combine two of these concepts:an exponentialrandom graph and a centrality index. Exponential random graphs comprise the most usefulclass of probability models for network data.These models often require the assumptionof a complex dependence structure, which creates certain difficulties in the estimation ofunknown model parameters. However, in the context of dynamic networks the exponentialrandom graph model provides the opportunity to incorporate a complex network structuresuch as centrality without the usual drawbacks associated with parameter estimation. Thethesis employs this idea by proposing probability models that are equivalent to the logisticregression models and that can be used to explain behaviour of both static and dynamicnetworks.