PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

A Uniform Lower Error Bound for Half-space Learning
Andreas Maurer and Massimiliano Pontil
In: Algorithmic Learning Theory, 13-16 Oct 2008, Budapest.

Abstract

We give a lower bound for the error of any unitarily invariant algorithm learning half-spaces against the uniform or related distributions on the unit sphere. The bound is uniform in the choice of the target half-space and has an exponentially decaying deviation probability in the sample. The technique of proof is related to a proof of the Johnson Lindenstrauss Lemma. We argue that, unlike previous lower bounds, our result is well suited to evaluate the benefits of multi-task or transfer learning, or other cases where an expense in the acquisition of domain knowledge has to be justified.

EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Learning/Statistics & Optimisation
Theory & Algorithms
ID Code:5356
Deposited By:Massimiliano Pontil
Deposited On:24 March 2009