English

Exact Post-Selection Inference for Sequential Regression Procedures

Methodology 2015-10-13 v5

Abstract

We propose new inference tools for forward stepwise regression, least angle regression, and the lasso. Assuming a Gaussian model for the observation vector y, we first describe a general scheme to perform valid inference after any selection event that can be characterized as y falling into a polyhedral set. This framework allows us to derive conditional (post-selection) hypothesis tests at any step of forward stepwise or least angle regression, or any step along the lasso regularization path, because, as it turns out, selection events for these procedures can be expressed as polyhedral constraints on y. The p-values associated with these tests are exactly uniform under the null distribution, in finite samples, yielding exact type I error control. The tests can also be inverted to produce confidence intervals for appropriate underlying regression parameters. The R package "selectiveInference", freely available on the CRAN repository, implements the new inference tools described in this paper.

Keywords

Cite

@article{arxiv.1401.3889,
  title  = {Exact Post-Selection Inference for Sequential Regression Procedures},
  author = {Ryan J. Tibshirani and Jonathan Taylor and Richard Lockhart and Robert Tibshirani},
  journal= {arXiv preprint arXiv:1401.3889},
  year   = {2015}
}

Comments

26 pages, 5 figures

R2 v1 2026-06-22T02:46:58.841Z