On classification tools for genetic algorithms

Some tools to measure convergence properties of genetic algorithms are introduced. A classification procedure is proposed for genetic algorithms based on a conjecture: the entropy and the fractal dimension of trajectories produced by them are quantities that characterize the classes of the algorithms. The role of these quantities as invariants of the algorithm classes is discussed together with the compression ratio of points of the genetic algorithm.

[1]  Y. Suhov,et al.  An entropic measurement of queueing behaviour in a class of manufacturing operations , 2001, Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.

[2]  W. Szlenk,et al.  An introduction to the theory of smooth dynamical systems , 1984 .

[3]  David H. Wolpert,et al.  No free lunch theorems for optimization , 1997, IEEE Trans. Evol. Comput..

[4]  Zbigniew Michalewicz,et al.  Genetic Algorithms + Data Structures = Evolution Programs , 1992, Artificial Intelligence.

[5]  R. Devaney,et al.  Chaos and Fractals: The Mathematics Behind the Computer Graphics , 1989 .

[6]  Zbigniew Michalewicz,et al.  Fractal Dimension of Trajectory as Invariant of Genetic Algorithms , 2006, ICAISC.

[7]  Thomas M. English.  No more lunch: analysis of sequential search , 2004, Proceedings of the 2004 Congress on Evolutionary Computation (IEEE Cat. No.04TH8753).

[8]  Michael D. Vose,et al.  Modeling Simple Genetic Algorithms , 1995, Evolutionary Computation.

[9]  D. Ornstein,et al.  On isomorphism of weak Bernoulli transformations , 1970 .

[10]  Michael D. Vose,et al.  Modeling Simple Genetic Algorithms , 1992, FOGA.

[11]  Zbigniew Michalewicz,et al.  Genetic Algorithms + Data Structures = Evolution Programs , 1996, Springer Berlin Heidelberg.

[12]  D. Ornstein.  Ergodic theory, randomness, and dynamical systems , 1974 .

[13]  Marc Toussaint,et al.  A No-Free-Lunch Theorem for Non-Uniform Distributions of Target Functions , 2004 .