Research-Complex likelihood ratio test for generalized linear models

Research

Probability & Statistics Seminar

Complex likelihood ratio test for generalized linear models

Speaker: Xingwei TONG (Beijing Normal University)
Time: Nov 18, 2018, 11:30-12:30
Location: Conference Room 415, Hui Yuan 3#

In this talk, we consider a special hypothesis test. First, we use it to analyze the common linear model and proposed a decision rule to control Type I error. Then we extend it to the generalized linear model (GLM). To be compatible with GLM, we propose a new definition with complex likelihood function and complete likelihood ratio test for GLM, which better fits the problem.