Testing and Estimation of Social Network Dependence with Time to Event Data
Abstract: Studying social network dependence is an emerging research area. In this talk, I will introduce two models for studying social network dependence with time-to-event data. First, I will present a latent Cox model with contextual e
ffects, which models the susceptibility of a user to the influence of his or her friends. Both testing and estimation of the social network dependence will be discussed. Second, I will present a class of conditional survival models for studying the contagion-based social network dependence. The associated nonparametric maximum likelihood estimation will be discussed. Simulations and applications to a mobile game data will be presented to demonstrate the performance of the proposed methods.