After an introduction that includes a discussion of the classic random walk, this paper presents a step-by-step development of the differential evolution (DE) global numerical optimization algorithm. Five fundamental DE strategies, each more complex than the last, are evaluated based on their conformance to invariance and symmetry principles, degree of control parameter dependence, computational efficiency and response to randomization. Optimal control parameter settings for the family of convex, quadratic functions are empirically derived. © Springer-Verlag Berlin Heidelberg 2013.
CITATION STYLE
Price, K. V. (2013). Differential Evolution. Intelligent Systems Reference Library, 38, 187–214. https://doi.org/10.1007/978-3-642-30504-7_8
Mendeley helps you to discover research relevant for your work.