snigdhagit / best-subset

Comparisons between best subset selection and other popular estimators for sparse regression

Best Subset Selection and Related Tools

Trevor Hastie, Rob Tibshirani, Ryan Tibshirani

Maintained by Ryan Tibshirani

This project contains an implementation of best subset selection in regression, based on a mixed integer quadratic program formulation of the subset selection problem and the Gurobi mixed integer program optimizer. It also contains tools for running simulations comparing best subset selection to other common sparse regression estimators such as the lasso and forward stepwise selection.

The mixed integer programming formulation of subset selection and simulation setup is based on the paper: Bertsimas, King, Mazumder (2016), "Best subset selection via a modern optimization lens".

For our discussion paper on best subset selection, forward stepwise, and the lasso, see: https://arxiv.org/abs/1707.08692.

Install the R package

To install the bestsubset R package directly from github, run the following in R:

library(devtools)
install_github(repo="ryantibs/best-subset", subdir="bestsubset")