Skip to main navigation Skip to search Skip to main content

Pac-bayesian aggregation of affine estimators

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Aggregating estimators using exponential weights depending on their risk appears optimal in expectation but not in probability. We use here a slight overpenalization to obtain oracle inequality in probability for such an explicit aggregation procedure. We focus on the fixed design regression framework and the aggregation of linear estimators and obtain results for a large family of linear estimators under a non-necessarily independent sub-Gaussian noise assumptions.

Original languageEnglish
Title of host publicationNonparametric Statistics- 3rd ISNPS 2016
EditorsPatrice Bertail, Delphine Blanke, Pierre-André Cornillon, Eric Matzner-Løber
PublisherSpringer New York LLC
Pages133-144
Number of pages12
ISBN (Print)9783319969404
DOIs
Publication statusPublished - 1 Jan 2018
Event3rd Conference of the International Society for Nonparametric Statistics, ISNPS 2016 - Avignon, France
Duration: 11 Jun 201616 Jun 2016

Publication series

NameSpringer Proceedings in Mathematics and Statistics
Volume250
ISSN (Print)2194-1009
ISSN (Electronic)2194-1017

Conference

Conference3rd Conference of the International Society for Nonparametric Statistics, ISNPS 2016
Country/TerritoryFrance
CityAvignon
Period11/06/1616/06/16

Fingerprint

Dive into the research topics of 'Pac-bayesian aggregation of affine estimators'. Together they form a unique fingerprint.

Cite this