Testing and Estimation of Social Network Dependence with Time-To-Event Data

• Activity Number: 174 - Dynamic Network Modeling
• Type: Contributed
• Date/Time: Monday, July 31, 2017 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Learning and Data Science
• Author(s): Lin Su* and Wenbin Lu and Rui Song and Danyang Huang
• Companies: North Carolina State University and North Carolina State University and NC State University and Renmin University
• Keywords: EM algorithm ; Latent spatial autocorrelation Cox model ; Social network dependence ; Time-to-event data

Abstract

Nowadays, events are spread rapidly along social network since we can share information with friends easily. We are interested in how people are aaffected by their friends' behavior. For example, if a person share the game he or she is playing, will his or her friends start playing it as well? Studying social network dependence is an emerging research area. In this work, we propose a novel latent spatial auto- correlation Cox model to study social network dependence with time-to-event data. The proposed model introduces a latent indicator to characterize whether a person might be aaffected by his or her friends' behavior. We first propose a score-type test for detecting the existence of social network dependence. If it exists, we further develop an EM-type algorithm to estimate the model parameters. The performance of the proposed test and estimators are illustrated by simulation studies and an application to a time-to-event data set about playing a popular QQ game from Tencent.