Translator Disclaimer
2015 High-dimensional inference in misspecified linear models
Peter Bühlmann, Sara van de Geer
Electron. J. Statist. 9(1): 1449-1473 (2015). DOI: 10.1214/15-EJS1041

Abstract

We consider high-dimensional inference when the assumed linear model is misspecified. We describe some correct interpretations and corresponding sufficient assumptions for valid asymptotic inference of the model parameters, which still have a useful meaning when the model is misspecified. We largely focus on the de-sparsified Lasso procedure but we also indicate some implications for (multiple) sample splitting techniques. In view of available methods and software, our results contribute to robustness considerations with respect to model misspecification.

Citation

Download Citation

Peter Bühlmann. Sara van de Geer. "High-dimensional inference in misspecified linear models." Electron. J. Statist. 9 (1) 1449 - 1473, 2015. https://doi.org/10.1214/15-EJS1041

Information

Received: 1 March 2015; Published: 2015
First available in Project Euclid: 7 July 2015

zbMATH: 1327.62420
MathSciNet: MR3367666
Digital Object Identifier: 10.1214/15-EJS1041

Subjects:
Primary: 62J07
Secondary: 62F25

Rights: Copyright © 2015 The Institute of Mathematical Statistics and the Bernoulli Society

JOURNAL ARTICLE
25 PAGES


SHARE
Vol.9 • No. 1 • 2015
Back to Top