When considering continuous search spaces, evolution strategies are among the well-performing metaheuristics. In contrast to other evolutionary algorithms, their main search operator is mutation which necessitates its adaptation during the run. Here, the covariance matrix plays an important role. Modern Evolution Strategies apply forms of covariance matrix adaptation. However, the quality of the common estimate of the covariance is known to be questionable for high search space dimensions. This paper presents a new approach by considering sparse covariance matrix techniques together with a space transformation.
展开▼
