Automatic Selection Pressure Control in Genetic Programming

Selection pressure must be dynamically managed in response to the changing evolutionary process in order to improve the effectiveness and efficiency of genetic programming (GP) systems using tournament selection. Instead of changing the tournament size and/or the population size via an arbitrary function to influence the selection pressure, this paper focuses on designing an automatic selection pressure control approach. In our approach, populations are clustered based on a dynamic program property. Then clusters become tournament candidates. The selection pressure in the tournament selection method is automatically changed during evolution according to the dynamically changing number of tournament candidates. Our approach is compared with the standard GP system (with no selection pressure control) on two problems with different kinds of fitness distributions. The results show that the automatic selection pressure control approach can improve the effectiveness and efficiency of GP systems

[1]  Vic Ciesielski,et al.  Prevention of early convergence in genetic programming by replacement of similar programs , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).

[2]  Lothar Thiele,et al.  A Comparison of Selection Schemes Used in Evolutionary Algorithms , 1996, Evolutionary Computation.

[3]  Sean Luke,et al.  Population Implosion in Genetic Programming , 2003, GECCO.

[4]  Mengjie Zhang,et al.  Population Clustering in Genetic Programming , 2006, EuroGP.

[5]  Leonardo Vanneschi,et al.  Dynamic Size Populations in Distributed Genetic Programming , 2005, EuroGP.

[6]  M. Tomassini,et al.  Saving computational effort in genetic programming by means of plagues , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[7]  David E. Goldberg,et al.  Genetic Algorithms, Tournament Selection, and the Effects of Noise , 1995, Complex Syst..

[8]  Huayang Xie,et al.  Diversity Control in GP with ADF for Regression Tasks , 2005, Australian Conference on Artificial Intelligence.

[9]  Steven M. Gustafson.  An analysis of diversity in genetic programming , 2004 .

[10]  Peter J. Angeline,et al.  Explicitly Defined Introns and Destructive Crossover in Genetic Programming , 1996 .

[11]  L. Darrell Whitley,et al.  Unbiased tournament selection , 2005, GECCO '05.

[12]  Mengjie Zhang,et al.  A Study of Good Predecessor Programs for Reducing Fitness Evaluation Cost in Genetic Programming , 2006, 2006 IEEE International Conference on Evolutionary Computation.

[13]  Leonardo Vanneschi,et al.  A new technique for dynamic size populations in genetic programming , 2004, Proceedings of the 2004 Congress on Evolutionary Computation (IEEE Cat. No.04TH8753).

[14]  John R. Koza,et al.  Genetic programming - on the programming of computers by means of natural selection , 1993, Complex adaptive systems.