CMA-ES

The CMA ( Covariance Matrix Adaptation) is a derandomisiertes method for adaptation of the covariance matrix of the Gaussian mutation distribution (normal distribution ) in the Evolution Strategy ( ES). The covariance matrix of the multidimensional ( multivariate ) normal distribution describes the pairwise dependencies between the variables. The adaptation of the covariance matrix in the evolution strategy is similar to the approximation of the inverse Hessian matrix in the classical optimization, such as the quasi -Newton method.

Basic principle

The adaptation principle is based on the idea to increase the probability of previously successful steps. To the covariance matrix of the distribution is changed such that increases the probability of the selected step of the last generation.