作者: G. Yin , G. Rudolph , H.-P, Schwefel
DOI: 10.1162/EVCO.1995.3.4.473
关键词: Approximation error 、 Evolutionary computation 、 Applied mathematics 、 Convergence (routing) 、 Rate of convergence 、 Evolution strategy 、 Stochastic approximation 、 Mathematical optimization 、 Mathematics 、 Dynamical systems theory 、 Iterated function
摘要: The main objective of this paper is to analyze the (1, λ) evolution strategy by use stochastic approximation methods. Both constant and decreasing step size algorithms are studied. Convergence estimation error bounds for developed. First algorithm converted a recursively defined scheme type. Then analysis carried out using analytic tools from approximation. In lieu examining discrete iterates, suitably scaled sequences defined. These interpolated then studied in detail. It shown that limits have natural connections certain continuous time dynamical systems.