Academic
Publications
Log-Linear Convergence and Divergence of the Scale-Invariant (1+1)ES in Noisy Environments

Log-Linear Convergence and Divergence of the Scale-Invariant (1+1)ES in Noisy Environments,10.1007/s00453-010-9403-3,Algorithmica,Mohamed Jebalia,Anne

Log-Linear Convergence and Divergence of the Scale-Invariant (1+1)ES in Noisy Environments   (Citations: 2)
BibTex | RIS | RefWorks Download
Noise is present in many real-world continuous optimization problems. Stochastic search algorithms such as Evolution Strategies (ESs) have been proposed as effective search methods in such contexts. In this paper, we provide a mathematical analysis of the convergence of a (1+1)-ES on unimodal spherical objective functions in the presence of noise. We prove for a multiplicative noise model that for a positive expected value of the noisy objective function, convergence or divergence happens depending on the infimum of the support of the noise. Moreover, we investigate convergence rates and show that log-linear convergence is preserved in presence of noise. This result is a strong theoretical foundation of the robustness of ESs with respect to noise.
Journal: Algorithmica , vol. 59, no. 3, pp. 425-460, 2011
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
Sort by: