PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Sharpening Occam's Razor
M. Li, J. Tromp and Paul M.B. Vitanyi
Information Processing Letters Volume 85, Number 5, pp. 267-274, 2003.

Abstract

We provide a new representation-independent formulation of Occam's razor theorem, based on Kolmogorov complexity. This new formulation allows us to: (i) Obtain better sample complexity than both length-based \cite{blumer1} and VC-based \cite{blumer} versions of Occam's razor theorem, in many applications; and (ii) Achieve a sharper reverse of Occam's razor theorem than that of \cite{board}. Specifically, we weaken the assumptions made in \cite{board} and extend the reverse to superpolynomial running times.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Learning/Statistics & Optimisation
Theory & Algorithms
ID Code:817
Deposited By:Paul Vitányi
Deposited On:01 January 2005