The out-of-bootstrap approach for model averaging and selection
J. Sunil Rao
Department of Biostatistics - Cleveland Clinic
We propose a bootstrap-based method for model averaging and selection that
focuses on training points that are left out of individual bootstrap
samples. This information can be used to estimate optimal weighting factors
for combining estimates from different bootstrap samples, and also
for finding the best subsets in the linear model setting. These proposals
provide alternatives to Bayesian approaches to model averaging and selection, re
quiring less computation and fewer subjective choices. In a sense,
we "enjoy the Bayesian omelette without making a mess in the kitchen".
This is joint work with Rob Tibshirani of the University of Toronto.