Initial Study on Handling Constrained Optimization Problems in Learnable Evolution Model

Date

2006-07

Authors

Wojtusiak, Janusz

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Learnable Evolution Model (LEM) is an evolutionary computation methodology that applies hypothesis formulation and instantiation to create new individuals. Initial study has shown that LEM significantly outperforms standard evolutionary computation methods in terms of evolution length on selected benchmark optimization problems. This paper presents initial results from handling constrained optimization problems in LEM. Constraints are classified as instantiable, which can be handled directly during instantiation process, and general, which cannot be directly instantiated. The latter can be handled by applying three different methods presented in this paper.

Description

Keywords

Concept learning, Induction, Constrained optimization

Citation

Wojtusiak, J., "Initial Study on Handling Constrained Optimization Problems in Learnable Evolution Model," Proceedings of The Graduate Student Workshop at Genetic and Evolutionary Computation Conference, GECCO 2006, Seattle, WA, July 8-12, 2006.