报告题目:用于假设检验和分类的一个两成分混合模型 (A two-component mixture model for hypothesis testing and classification)
报告人:吴静静教授 加拿大卡尔加里大学(Jingjing Wu, University of Calgary)
邀请人:赵强副教授
报告摘要:In this research, we studied a two-component nonparametric mixture model with stochastic dominance constraint, a model arises naturally from genetic studies. Our interest lies in both the estimation of mixing proportion and classification. For this model, we proposed and studies a nonparametric estimation based on cumulative distribution functions and a maximum likelihood estimation (MLE) through multinomial approximation. In order to incorporate nicely the stochastic dominance constraint, we introduced a semiparametric model for which we proposed and investigated both MLE and minimum Hellinger distance estimation (MHDE). We also proposed a hypothesis testing to test the validity of the semiparametric model. For the proposed methods, we investigated both their asymptotic properties such as consistency and asymptotic normality and their finite-sample performance through simulation studies and real data analysis.