A bootstrap recipe for post-model-selection inference under linear regression models

• 成果类型: Article
• 署名作者: Lee, S. M. S.; Wu, Y.
• 署名单位: University of Hong Kong; University of Waterloo
• 刊物名称: BIOMETRIKA
• ISSN/ISSBN: 0006-3444
• DOI: 10.1093/biomet/asy046
• 发表日期: 2018
• 页码: 873890
• 关键词: least-squares; confidence-intervals; variable selection; Lasso

摘要

We propose a general bootstrap recipe for estimating the distributions of post-model-selection least squares estimators under a linear regression model. The recipe constrains residual bootstrapping within the most parsimonious, approximately correct, models to yield a distribution estimator which is consistent provided any wrong candidate model is sufficiently separated from the approximately correct ones. Our theory applies to a broad class of model selection methods based on information criteria or sparse estimation. The empirical performance of our procedure is illustrated with simulated data.