Robust Linear Model Selection Based on Bootstrap Robust LARS



We provide R code to sequence candidate predictors using the bootstrap robust LARS procedure. An example illustrates the use of this code.
Questions can be sent to Stefan Van Aelst.

R code

Main function to sequence candidate predictors using the bootstrap robust LARS procedure: B-RLARS.R
An example that illustrates the use of the above code: EXAMPLE.R

Data sets analyzed in the paper listed below

US demographic data
Structure of the variables in the protein data set: Description
Complete protein training data set (compressed, 22MB): Protein data
Reduced protein data set: Training data Test data

Reference

Khan, J.A., Van Aelst, S., and Zamar, R.H. (2007), " Robust Linear Model Selection Based on Least Angle Regression, " Journal of the American Statistical Association , 102, 1289-1299.
Paper