The Annals of Statistics
- Ann. Statist.
- Volume 47, Number 3 (2019), 1475-1504.
Valid confidence intervals for post-model-selection predictors
We consider inference post-model-selection in linear regression. In this setting, Berk et al. [Ann. Statist. 41 (2013a) 802–837] recently introduced a class of confidence sets, the so-called PoSI intervals, that cover a certain nonstandard quantity of interest with a user-specified minimal coverage probability, irrespective of the model selection procedure that is being used. In this paper, we generalize the PoSI intervals to confidence intervals for post-model-selection predictors.
Ann. Statist., Volume 47, Number 3 (2019), 1475-1504.
Received: January 2016
Revised: May 2018
First available in Project Euclid: 13 February 2019
Permanent link to this document
Digital Object Identifier
Mathematical Reviews number (MathSciNet)
Zentralblatt MATH identifier
Bachoc, François; Leeb, Hannes; Pötscher, Benedikt M. Valid confidence intervals for post-model-selection predictors. Ann. Statist. 47 (2019), no. 3, 1475--1504. doi:10.1214/18-AOS1721. https://projecteuclid.org/euclid.aos/1550026846
- Appendix: Proofs, algorithms, comments, details and extensions. The Appendix contains the following material: comments on the assumptions made on the error variance; proofs of the results given in Sections 2 and 3; additional material for Sections 2 and 3; descriptions of the algorithms for computing the PoSI confidence intervals; details concerning the numerical calculations for Section 4; additional simulation results.