Add to Chrome
Add to Firefox
Add to EdgeOn the Robustness of Median Sampling in Noisy Evolutionary Optimization
Jul 28, 2019
In real-world optimization tasks, the objective (i.e., fitness) function evaluation is often disturbed by noise due to a wide range of uncertainties. Evolutionary algorithms (EAs) have been widely applied to tackle noisy optimization, where reducing the negative effect of noise is a crucial issue. One popular strategy to cope with noise is sampling, which evaluates the fitness multiple times and uses the sample average to approximate the true fitness. In this paper, we introduce median sampling as a noise handling strategy into EAs, which uses the median of the multiple evaluations to approximate the true fitness instead of the mean. We theoretically show that median sampling can reduce the expected running time of EAs from exponential to polynomial by considering the (1+1)-EA on OneMax under the commonly used one-bit noise. We also compare mean sampling with median sampling by considering two specific noise models, suggesting that when the 2-quantile of the noisy fitness increases with the true fitness, median sampling can be a better choice. The results provide us with some guidance to employ median sampling efficiently in practice.
