
12
Bayesian Variable Selection for Predictively
Optimal Re gression
Tanujit Dey
The College of William and Mary
Ernest Fokoué
Rochester Institute of Technology
CONTENTS
12.1 Intro ductio n . . . . . . . . .. . . . . . .. .. . . . . . .. . . .. . . . .. . . .. . . . .. .. . . . .. .. . 245
12.2 Appro aches to Response Prediction in Reg ression . . . . . . . . .. . . .. 247
12.2.1 Basic elements a nd assumptions .. . . . . . . . .. . . . . . .. .. . . . . 247
12.2.2 Methods of response prediction . . .. .. . . . . . .. . . .. . . . .. .. . 250
12.3 Bayesian Approaches to Variable Selection . . . . . . . .. . . . . . . . .. . . . . 253
12.3.1 Bayesian varia ...