View article

Evolution strategies are a class of evolutionary algorithms for black-box optimization and achieve state-of-the-art performance on many benchmarks and real-world applications. Evolution strategies typically evolve a Gaussian distribution to approach the optimum. In this paper, we present a survey of recent advances in evolution strategies. We summarize the techniques, extensions, and practical considerations of evolution strategies for various optimization problems. We discuss some important open questions and promising topics that desire further research. Many of the discussed techniques and principles are applicable to other algorithms.