Skip navigation
Skip navigation

Obtaining fast error rates in nonconvex situations

Mendelson, Shahar

Description

We show that under mild assumptions on the learning problem, one can obtain a fast error rate for every reasonable fixed target function even if the base class is not convex. To that end, we show that in such cases the excess loss class satisfies a Bernstein type condition.

CollectionsANU Research Publications
Date published: 2008
Type: Journal article
URI: http://hdl.handle.net/1885/39068
Source: Journal of Complexity
DOI: 10.1016/j.jco.2007.09.001

Download

File Description SizeFormat Image
01_Mendelson_Obtaining_fast_error_rates_in_2008.pdf264.15 kBAdobe PDF    Request a copy


Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.

Updated:  12 November 2018/ Responsible Officer:  University Librarian/ Page Contact:  Library Systems & Web Coordinator