报告人简介:
赵晓兵,博士,现为浙江财经大学数据科学学院教授,博士生导师。主要研究领域包括治愈模型、复发事件模型、面板计数数据、高维数据降维、大规模网络数据分析等。先后主持国家自然科学基金面上项目、国家社科基金一般项目等国家级项目4项,在主流统计学和精算学期刊,例如 Statistics in Medicine; Statistica Sinica; Journal of Multivariate Analysis, Journal of Statistical Planning and Inference ; Computational Statistics and Data Analysis; Lifetime Data Analysis; Insurance: Mathematics and Economics等发表论文60多篇。
报告摘要:
The analysis of dynamic network data based on statistical models has attracted considerable attention in social and biological research fields. In this paper, we propose a statistical model for the recurrent events of instantaneous interactions between the nodes using a Poisson process with a semiparametric mean function of the recurrent interactions and latent membership of the nodes. A joint model of the recurrent interactions process and a discrete-time observation process is proposed to characterize the impact of the time-slices on the snapshots. A variational expectation-maximization algorithm is applied to estimate the connectivity parameters and the latent variables. Simulation studies and real data are used to illustrate the performance of the proposed model and methodology.