國立東華大學應用數學系

專    題    演    講

主講人：Prof. Wei-Wen Hsu

                 Department of Statistics, Kansas State University.

講  題：A Robust score test of homogeneity for Zero-Inflated count data    

時  間：106 年 05月 31日 (星期三)  10:00-12:00

地  點：理工一館 A324會議室

摘 要

Score statistics are often used for the test of homogeneity in zero-inflated models for count data. The most cited justification is that it only requires the model fit under the null hypothesis. However, the true null models are often unknown in practice and these statistics can be invalid when the null models are not correctly specified. As an evidence, an intensive simulation study was conducted and revealed that the empirical sizes of these score tests for zero-inflated models might behave extremely liberal and unstable under misspecifications, particularly when the mean function of baseline distribution or the baseline distribution itself was misspecified. In this study, we propose a new score test of homogeneity for zero-inflated models which is robust to these misspecifications. Technically, the test is developed under the framework of Poisson-Gamma mixture model which can provide a more general framework to incorporate various baseline distributions without specifying the mean function. The empirical performances of the proposed test in finite samples are evaluated in simulation studies and the dental caries data from the Detroit Dental Health Project (DDHP) and Girl Scout data from Scouting Nutrition and Activity Program (SNAP) are used to illustrate the proposed test.

※※※   歡      迎      參      加   ※※※se1060531 附檔：