EVENTS
Following is a more detailed description of the event. Please note that not all events are open to the general public. If you need additional information relating to an event, please communicate with the contact person for the event or email www@isye.gatech.edu.
STATISTICS SERIES :: Tuning Variable Selection Procedures
GUEST LECTURER
Dennis Boos
AFFILIATION
North Carolina State University
ABSTRACT
Two new approaches to variable selection in regression are presented. The key idea in both approaches is to calibrate an existing tunable selection method in order to achieve desirable properties of selected models. Attention is restricted to forward selection for which the tuning parameter is the familiar alpha-to-enter. In the first approach, Noise Added Model Selection (NAMS), parametric bootstrap-like data sets are generated by incrementally adding noise to the response variable, and alpha is tuned by tracking the effect of added noise on selected models' mean squared errors for different alpha values. In the second approach, Variable Added Model Selection (VAMS), random phony predictor variables are added to the data set, and alpha is tuned by tracking the proportion of falsely included phony variables in the models selected for different alpha values.
DATE & TIME
Thursday, October 12, 2006 -- 11:00 AM
DURATION
Not Specified
LOCATION
Executive Classroom (ISyE Main Building)
CONTACT PERSON
yajun DOT mei AT isye DOT gatech DOT edu
