New Hybrid Variants of Genetic Algorithms

Many problems of combinatorial optimization belong to the class of NP-complete problems and can be solved efficiently only by heuristics like Genetic Algorithms or Evolution Strategies. However, both methods still have a number of drawbacks that reduce their ability to solve that kind of problems. To increase solution quality plenty of work has been investigated in order to introduce new coding standards and operators especially for Genetic Algorithms. Almost all of these approaches have one thing in common: They are very problem specific and often do not even consider the basic principles of Genetic Algorithms properly. In this monograph we take a different approach and look upon the concept of Standard Genetic Algorithms as an artificial self organizing process in order to overcome some of the fundamental problems. With the purpose of making the algorithm more open for scalability on the one hand, and retarding premature convergence on the other hand, this paper presents new extensions of the Standard Genetic Algorithm that are not based on any problem spezific knowledge. The experimental part of the thesis shows the new algorithms superior performance when solving multimodal combinatorial optimization problems.

Produktdetails

• Autoren: Michael Affenzeller
• Verlag: Trauner
• Sprache: Englisch
• Produktform: Taschenbuch
• Erschienen: 01.01.2003
• EAN: 9783854874850
• ISBN: 978-3-85487-485-0
• Seiten: 150
• Gewicht: 205 g
• Serien: Schriften der Johannes-Kepler-Universität Linz, Reihe C; Schriften der Johannes-Kepler-Universität Linz, Reihe C
• Thema: Naturwissenschaften, Medizin, Informatik, Technik > Mathematik > Wahrscheinlichkeitstheorie, Stochastik, Mathematische Statistik