TitleAn Evaporation Mechanism for Dynamic and Noisy Multimodal Optimization
Publication TypeConference Paper
Year of Publication2009
AuthorsFernández JLuis, Arcos JLluis
Conference NameProceedings of the 10th Genetic and Evolutionary Computation Conference (GECCO)
Pagination17-24
Abstract

Dealing with imprecise information is a common characteristic in real-world problems. Specifically, when the source of the information are physical sensors, a level of noise in the evaluation has to be assumed. Particle Swarm Optimization is a technique that presented a good behavior when dealing with noisy fitness functions. Nevertheless, the problem is still open. In this paper we propose the use of the evaporation mechanism for managing with dynamic multi-modal optimization problems that are subject to noisy fitness functions. We will show how the evaporation mechanism does not require the detection of environment changes and how can be used for improving the performance of PSO algorithms working in noisy environments.