Adaptive Encoding for Optimization

This report describes a general method for rendering search coordinate system independent, Adaptive Encoding (AE). Adaptive Encoding is applicable to any continuous domain search algorithm and includes (incremental) changes of the coordinate system, that is, changes of the representation of solutions. One attractive way to change the representation within AE is derived from the Covariance Matrix Adaptation (CMA). We prove that adaptive encoding recovers the CMA Evolution Strategy, when suitably applied to an evolution strategy with cumulative step-size adaptation. The proof implies that adaptive encoding provides the means to apply CMA-like representation changes to any search algorithm in continuous domain. Key-words: optimization, evolutionary algorithms, encoding, representation, coordinate system, covariance matrix adaptation, invariance ∗ Adaptive Combinatorial Search Team, Microsoft Research–INRIA Joint Centre. 28, rue Jean Rostand, 91893 Orsay Cedex, France. email:[email protected]. Encodage adaptif pour optimisation Résumé : Pas de résumé Mots-clés : Pas de motclef Adaptive Encoding for Optimization 3