主 题：Cross Validation for Selecting a Parametric/Nonparametric Regression Procedure
主讲人：Prof. Yuhong Yang
主持人：林华珍教授
时 间：2015年7月14日下午4点-5点
地 点：通博楼B212学术会议室
主办单位：统计研究中心 统计学院 科研处
主讲人简介：
Yuhong Yang obtained his Ph.D. at Yale in 1996 and then worked at Iowa State University as assistant and associate professor. Since 2007, he is professor at School of Statistics at the University of Minnesota. Yuhong's research interests include model selection, model averaging, high-dimensional data analysis, classification, and forecasting.Yuhong was a recipient of the US NSF CAREER AWARD. He is a fellow of the Institute of Mathematical Statistics. He served several journals as associate editor, including Annals of Statistics, Annals of Institute of Statistical Mathematics, Journal of Statistical Planning and Inference, Statistica Sinica, and Statistical Survey.
内容提要：
While there are various parametric and nonparametric procedures (with possible model selection) for high-dimensional regression, an unanswered but important question is how to select one of them for data at hand consistently for prediction or inference. The difficulty is due to that the targeted behaviors of the different procedures depend heavily on uncheckable or difficult-to-check assumptions on the data generating process. Fortunately, cross-validation (CV) provides a general tool to solve this problem. In this work, results are provided on how to apply CV to consistently choose the best method, yielding new insights and guidance for potentially vast amount of application. In addition, we address several seemingly widely spread misconceptions on CV. （Joint work with Yongli Zhang.)
