题目：Estimation of the Degree parameters in Directed Network Models and its Applications
主讲：晏挺
时间：2017年12月15日 15:00-16:00
地点：明德主楼1030会议室
摘要：The heterogeneity of degrees is common in network data. To characterize this phenomenon, network models often contain the degree parameter. Since each node is assigned at least one degree parameter, the dimension of the space of parameters increases as the number of nodes grows. It makes inference difficult. In this talk, I will present a unified theoretical framework on how to estimate the degree parameter in directed network models and establish the asymptotic theories of the estimator. The results are illustrated by three important applications to the probit model with reciprocity parameter, differential private bi-degree based models and the degree-corrected stochastic block model.
简介：
晏挺，中国科学技术大学统计学学士学位、博士学位，美国乔治华盛顿大学统计系博士后，华中师范大学统计系教授，
主要研究：Network models, Random Graphs, Paired comparisons
多个科研项目主持人：
多项国家自然科学基金面上项目：网络数据隐私保护的统计方法研究
国家自然科学基金主任基金:高维网络数据建模及其渐近推断.
国家自然科学基金青年基金: 加权网络数据建模及其统计推断. 3. 中央高校基本科研业务费专项资金探索创新项目：网络数据隐私保护及其统计推断研究
中央高校基本科研业务费专项资金探索创新项目：网络模型中的假设检验