主 题：Sufficient dimension reduction via distance covariance
主讲人：Xiangrong Yin教授
主持人：林华珍教授
时 间：2014年6月23日（星期一）下午4:30-5:30
地 点：通博楼B座212学术会议室
主办单位：统计学院 统计研究中心 科研处
主讲人简介：
Xiangrong Yin，The University of Georgia 教授，2001荣获 The Inaugural Editor’s Award for the best article published in the Australian and New Zealand，累计资助236,577.51美元,是International Chinese Statistical Association（ICSA）、IMS、ASA会员，是期刊Journal of Nonparametric Statistic和Statistics and Probability Letters的 Associate Editor，发表论文40余篇，包括发表在统计学顶级期刊Annals of Statistics及Biometrics上的论文多篇。
内容提要：
We introduce a novel approach to sufficient dimension reduction problems using distance covariance. Our method requires very mild conditions on the predictors. It estimates the central subspace efficiently even when many pre-dictors are categorical or discrete. Our method keeps the model-free advantage without estimating link function. Under regularity conditions, root-n consis-tency and asymptotic normality are established for our estimator. We compare the performance of our method with some existing dimension reduction meth-ods by simulations and find that our method is very competitive and robust across a number of models. We also analyze the Australian Institute of Sport data to demonstrate the efficacy of our method.
