Adaptive interactive genetic algorithms with individual interval fitness

Abstract It is necessary to enhance the performance of interactive genetic algorithms in order to apply them to complicated optimization problems successfully. An adaptive interactive genetic algorithm with individual interval fitness is proposed in this paper in which an individual fitness is expressed by an interval. Through analyzing the fitness, information reflecting the distribution of an evolutionary population is picked up, namely, the difference of evaluating superior individuals and the difference of evaluating a population. Based on these, the adaptive probabilities of crossover and mutation operators of an individual are presented. The algorithm proposed in this paper is applied to a fashion evolutionary design system, and the results show that it can find many satisfactory solutions per generation. The achievement of the paper provides a new approach to enhance the performance of interactive genetic algorithms.

[1]  Hideyuki Takagi,et al.  Discrete fitness values for improving the human interface in an interactive GA , 1996, Proceedings of IEEE International Conference on Evolutionary Computation.

[2]  Sung-Bae Cho,et al.  Application of interactive genetic algorithm to fashion design , 2000 .

[3]  Yong Zhou,et al.  Interactive genetic algorithms with multi-population adaptive hierarchy and their application in fashion design , 2007, Appl. Math. Comput..

[4]  Peter G. Anderson,et al.  Neural network fitness functions for a musical IGA , 1996 .

[5]  R. Dawkins.  The Blind Watchmaker , 1986 .

[6]  Joshua R. Smith.  Designing Biomorphs with an Interactive Genetic Algorithm , 1991, ICGA.

[7]  Wang Xu-fa.  Improvement on Convergence and Quality of User Evaluation in Interactive Genetic Algorithms , 2002 .

[8]  Hideyuki Takagi,et al.  Interactive evolutionary computation: fusion of the capabilities of EC optimization and human evaluation , 2001, Proc. IEEE.

[9]  Nao and Iba Hitoshi Tokui,et al.  Music Composition with Interactive Evolutionary Computation , 2000 .

[10]  Hideyuki Takagi,et al.  User Fatigue Reduction by an Absolute Rating Data-trained Predictor in IEC , 2006, 2006 IEEE International Conference on Evolutionary Computation.

[11]  Craig Caldwell,et al.  Tracking a Criminal Suspect Through "Face-Space" with a Genetic Algorithm , 1991, ICGA.

[12]  Xavier Llorà,et al.  Combating user fatigue in iGAs: partial ordering, support vector machines, and synthetic fitness , 2005, GECCO '05.