Omid Askarisichani successfully passed his PhD Thesis Proposal on the topic of "Explainable Models of Influence and Performance on Dynamic Networks".
Committee: Ambuj Singh (Chair), Francesco Bullo, Noah E. Friedkin, Xifeng Yan
Networks model complex systems in many real applications, including social media, finance, and political systems. In such settings, nodes often represent people, artificial agents, or political parties while edges portray their signed relationships. In most applications, the relationships modeled change over time. Often times, there is a need to estimate the underlying relationships and forecast their changes. Our research is at the intersection of machine learning, network science, and social science. In our studies, we use graph theory, NLP, and convex optimization to extract information on how to improve the performance of individuals in financial and social systems. We also study the reasons behind changes and shocks in networks.
By leveraging data, we study how the patterns of influence and relationships may impact the performance of stock traders. We build upon theories from sociology, namely structural balance theory—which describes the dynamics that govern the sentiment of interpersonal relationships—and assess the impact on stock traders' profitability. Moreover, we show a generalization of structural balance theory that describes the dynamics of relationships among countries over more than two decades. We capture their dynamics using a time-varying Markov model, and prove the convergence rate for the proposed model. Furthermore, we find the factors leading to an individual becoming influential in the underlying social system and efficiently estimate an individual's influence on others on the basis of their expertise, communication contents, and interaction patterns. We approach this estimation problem with convex optimization and neural network models.