2008 | OriginalPaper | Buchkapitel
Dynamic Adaptation of Genetic Operators’ Probabilities
verfasst von : Fatemeh Vafaee, Peter C. Nelson, Chi Zhou, Weimin Xiao
Erschienen in: Nature Inspired Cooperative Strategies for Optimization (NICSO 2007)
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
We propose a new method of dynamically adapting the probabilities of genetic operators based on the global behavior of the population for each generation. The proposed method consists of two main components which are assigning credits to operators according to the fitness improvements of the individuals, and updating the operators’ probabilities at the onset of each generation. Each of these components can be implemented based on various mathematical approaches; hitherto, two different variants have been investigated. To leverage our previous work we used Gene Expression Programming (GEP) as a benchmark to investigate the power of our novel approach. Nevertheless, this new method can be easily extended to other genetic programming variants. Our experimental results on two symbolic regression problems show that this method follows a faster convergence curve and it improves the performance considerably while imposing an insignificant additional cost.