PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Optimal robust expensive optimization is tractable
Philippe Rolet, Olivier Teytaud and Michele Sebag
GECCO '09: Proceedings of the 11th Annual conference on Genetic and evolutionary computation 2009.

Abstract

Following a number of recent papers investigating the possi- bility of optimal comparison-based optimization algorithms for a given distribution of probability on fitness functions, we (i) discuss the comparison-based constraints (ii) choose a setting in which theoretical tight bounds are known (iii) develop a careful implementation using billiard algorithms, Upper Confidence trees and (iv) experimentally test the tractability of the approach. The results, on still very simple cases, show that the approach, yet still preliminary, could be tested successfully until dimension 10 and horizon 50 it- erations within a few hours on a standard computer, with convergence rate far better than the best algorithms.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
Theory & Algorithms
ID Code:5880
Deposited By:Philippe Rolet
Deposited On:08 March 2010