On replacement strategies in steady state evolutionary algorithms
Smith, J. (2007) On replacement strategies in steady state evolutionary algorithms. Evolutionary Computation, 15 (1). pp. 29-59. ISSN 1530-9304
Publisher's URL: http://dx.doi.org/10.1162/evco.2007.15.1.29
This is one of a series of papers [e.g. proceedings of FOGA 2001, 2003, Gecco 2001] developing theoretical models of different aspects of complex adaptive systems and exploring their pragmatic utility. The paper provides an analytical synthesis of various different models of the evolution of fitness distributions in a general class of Evolutionary Algorithms. After deriving and comparing the analytical models for different algorithms, it examines their significance as predictors for algorithms' relative optimisation performance. It is shown that certain analytic indicators have little utility, whereas others, developed for the first time here, have significant impact for the designers requiring consistent behaviour.
Repository Staff Only: item control page
Total Document Downloads in Past 12 Months
Total Document DownloadsMore statistics for this item...