English

Quasi-Likelihood and/or Robust Estimation in High Dimensions

Statistics Theory 2013-01-07 v2 Methodology Statistics Theory

Abstract

We consider the theory for the high-dimensional generalized linear model with the Lasso. After a short review on theoretical results in literature, we present an extension of the oracle results to the case of quasi-likelihood loss. We prove bounds for the prediction error and 1\ell_1-error. The results are derived under fourth moment conditions on the error distribution. The case of robust loss is also given. We moreover show that under an irrepresentable condition, the 1\ell_1-penalized quasi-likelihood estimator has no false positives.

Keywords

Cite

@article{arxiv.1206.6721,
  title  = {Quasi-Likelihood and/or Robust Estimation in High Dimensions},
  author = {Sara van de Geer and Patric Müller},
  journal= {arXiv preprint arXiv:1206.6721},
  year   = {2013}
}

Comments

Published in at http://dx.doi.org/10.1214/12-STS397 the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat.org)

R2 v1 2026-06-21T21:27:30.829Z