PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Continuous lunches are free
Olivier Teytaud and Anne Auger
In: GECCO 2007, 2007, London.

Abstract

This paper investigates extensions of No Free Lunch (NFL) theorems to countably infinite and uncountable infinite do- mains. The original NFL due to Wolpert and Macready states that all search heuristics have the same performance when averaged over the uniform distribution over all possible functions. For infinite domains the extension of the concept of distribution over all possible functions involves measur- ability issues and stochastic process theory. For countably infinite domains, we prove that the natural extension of NFL theorems does not hold, but that a weaker form of NFL does hold, by stating the existence of non-trivial distributions of fitness leading to equal performance for all search heuristics. Our main result is that for continuous domains, NFL does not hold.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:3193
Deposited By:Olivier Teytaud
Deposited On:20 January 2008