A robust black-box optimization algorithm that adapts the covariance matrix of a search distribution to efficiently explore high-dimensional spaces.