σ-Self-Adaptive Weighted Multirecombination Evolution Strategy with Scaled Weights on the Noisy Sphere

Authors:
Hans-Georg Beyer;Alexander Melkozerov
Affiliations:
Research Center Process and Product Engineering Department of Computer Science, Vorarlberg University of Applied Sciences, Dornbirn, Austria A-6850;Research Center Process and Product Engineering Department of Computer Science, Vorarlberg University of Applied Sciences, Dornbirn, Austria A-6850
Venue:
Proceedings of the 10th international conference on Parallel Problem Solving from Nature: PPSN X
Year:
2008

Citing 6
Cited 0

The theory of evolution strategies

The theory of evolution strategies
Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (CMA-ES)

Evolutionary Computation
Completely Derandomized Self-Adaptation in Evolution Strategies

Evolutionary Computation
Weighted multirecombination evolution strategies

Theoretical Computer Science - Foundations of genetic algorithms
Mutative σ-self-adaptation can beat cumulative step size adaptation when using weighted recombination

Proceedings of the 10th annual conference on Genetic and evolutionary computation
Covariance Matrix Adaptation Revisited --- The CMSA Evolution Strategy ---

Proceedings of the 10th international conference on Parallel Problem Solving from Nature: PPSN X

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a performance analysis of the recentlyproposed σ-self-adaptive weighted recombinationevolution strategy (ES) with scaled weights. The steady statebehavior of this ES is investigated for the non-noisy and noisycase, and formulas for the optimal choice of the learning parameterare derived allowing the strategy to reach maximal performance. Acomparison between weighted multirecombination ES withσ-self-adaptation (σSA) and withcumulative step size adaptation (CSA) shows that the self-adaptiveES is able to reach similar (or even better) performance as its CSAcounterpart on the noisy sphere.