Spatial Variable Selection and An Application to Virginia Lyme Disease Emergence


主讲人:Hong Yili,Associate Professor of Virginia Tech, USA




主讲人介绍:Yili Hong received a BS in statistics in 2004 from University of Science and  Technology of China. He received his MS in statistics in 2005 and PhD in  statistics in 2009 from Iowa State University. He is currently an Associate  Professor in the Department of Statistics at Virginia Tech. His research mainly  focuses on machine learning and engineering applications, statistical  reliability, survival analysis and spatial data analysis. His research has been  published in top journals such as Technometrics, JQT, Annals of Applied  Statistics, JASA, IEEE Transactions on Reliability, and Quality Engineering. He  is one of the recipients of the 2011 DuPont Young Professor Award. He is an  associate editor for Technometrics and JQT. He is a co-guest editor for a  special issue on big data in reliability for JQT. He is an elected member of  International Statistical Institute.

内容介绍:We use a spatial Poisson regression model to link the spatial disease counts and  environmental and economic variables, and develop a spatial variable selection  procedure to effectively identify important factors by using an adaptive elastic  net penalty. The proposed methods can automatically select important covariates,  while adjusting for possible spatial correlations of disease counts. The  performance of the proposed method is studied and compared with existing methods  via a comprehensive simulation study. We apply the developed variable selection  methods to the Virginia Lyme disease data and identify important variables that  are new to the literature.