发布时间:2018-06-11
报告人:林巍(Ohio大学数学系)
报告题目:Applications of Influence Function in Dimension Reduction
报告摘要:Sufficient dimension reduction (SDR) models in a regression analysis assumes that the response variable depends on the predictor variables $X$ only through a few linear combinations $BX$. It sees much progress and becomes very popular with the introduction of the inverse regression method pioneered by Li (1991). Various methods have been proposed to analyze the SDR models and majority of them are central-matrix based. However, not much has been done regarding how these methods compare to each other or how to choose from them in practice. In this talk, we introduce how one can apply the influence function in the area of dimension reduction to help selecting an appropriate central matrix and to help identify the structural dimension of the SDR model.
报告时间:2018年6月15日(星期五)上午9:00 -10:00.
报告地点:科技楼南楼602