Report time: On May 11, 2016 (Wednesday) 2:30p.m.

Report place: The third conference room in Jingshi Academic Hall

Reporter: Junlong Zhao

Abstract：In a functional data regression model, screening of important variable is an important problem when there are multiple functional predictive variables. Outliers are often exist in practical issue，so that we should consider a robust method to choose variables. This paper bases on a robust method of LAD loss and functional principal component to choose variables. And then this paper studies asymptotic property of estimations and have some cases study.