English

Post-selection inference for L1-penalized likelihood models

Methodology 2016-10-17 v3

Abstract

We present a new method for post-selection inference for L1 (lasso)-penalized likelihood models, including generalized regression models. Our approach generalizes the post-selection framework presented in Lee et al (2014). The method provides p-values and confidence intervals that are asymptotically valid, conditional on the inherent selection done by the lasso. We present applications of this work to (regularized) logistic regression, Cox's proportional hazards model and the graphical lasso.

Keywords

Cite

@article{arxiv.1602.07358,
  title  = {Post-selection inference for L1-penalized likelihood models},
  author = {Jonathan Taylor and Robert Tibshirani},
  journal= {arXiv preprint arXiv:1602.07358},
  year   = {2016}
}

Comments

23 pages, 8 figures

R2 v1 2026-06-22T12:56:27.716Z