Abstract
Evolutionary Algorithms' (EAs') application to real world optimization problems often involves expensive fitness function evaluation. Naturally this has a crippling effect on the performance of any population based search technique such as EA. Estimating the fitness of individuals instead of actually evaluating them is a workable approach to deal with this situation. Optimization problems in real world often involve expensive fitness. In [14] and [15] we presented two EA models, namely DAFHEA (Dynamic Approximate Fitness based Hybrid Evolutionary Algorithm) and DAFHEA-II respectively. The original DAFHEA framework [14] reduces computation time by controlled use of meta-models generated by Support Vector Machine regression to partly replace actual fitness evaluation by estimation. DAFHEA-II [15] is an enhancement to the original framework in that it can be applied to problems that involve uncertainty. DAFHEA-II, incorporates a multiple-model based learning approach for the support vector machine approximator to filter out effects of noise [15]. In this paper we present further investigation on the performance of DAFHEA and DAFHEA-II.
Original language | English |
---|---|
Title of host publication | GECCO 2008 |
Place of Publication | USA |
Publisher | ACM Press |
Pages | 2117-2122 |
Number of pages | 6 |
ISBN (Electronic) | 9781605581316 |
Publication status | Published - 2008 |
Event | Genetic and Evolutionary Computation Conference - USA, New Zealand Duration: 12 Jul 2008 → 16 Jul 2008 |
Conference
Conference | Genetic and Evolutionary Computation Conference |
---|---|
Country/Territory | New Zealand |
Period | 12/07/08 → 16/07/08 |