Modeling GA Performance for Control Parameter Optimization

The optimization of the control parameters of genetic algorithms is often a time consuming and tedious task. In this work we take the meta-level genetic algorithm approach to control parameter optimization. We enhance this process by incorporating a neural network for fitness evaluation. This neural network is trained to learn the complex interactions of the genetic algorithm control parameters and is used to predict the performance of the genetic algorithm relative to values of these control parameters. To validate our approach we describe a genetic algorithm for the largest common subgraph problem that we develop using this neural network enhanced meta-level genetic algorithm. The resulting genetic algorithm significantly outperforms a hand-tuned variant and is shown to be competitive with a hill-climbing algorithm used in practical applications.

[1]  A. Shoukry,et al.  Neural network approach for solving the maximal common subgraph problem , 1996, IEEE Trans. Syst. Man Cybern. Part B.

[2]  James A. Hendler,et al.  Massively parallel matching of knowledge structures , 1994 .

[3]  K. Dejong,et al.  An analysis of the behavior of a class of genetic adaptive systems , 1975 .

[4]  Rajarshi Das,et al.  A Study of Control Parameters Affecting Online Performance of Genetic Algorithms for Function Optimization , 1989, ICGA.

[5]  Ching Y. Suen,et al.  Hierarchical attributed graph representation and recognition of handwritten chinese characters , 1991, Pattern Recognit..

[6]  S. Wu,et al.  GENETIC ALGORITHMS FOR NONLINEAR MIXED DISCRETE-INTEGER OPTIMIZATION PROBLEMS VIA META-GENETIC PARAMETER OPTIMIZATION , 1995 .

[7]  Horst Bunke,et al.  Fast Computation of Error-Correcting Graph Isomorphisms Based on Model Precompilation , 1997, ICIAP.

[8]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[9]  K. De Jong.  Adaptive System Design: A Genetic Approach , 1980, IEEE Transactions on Systems, Man, and Cybernetics.

[10]  William C. Regli,et al.  Resolving non-uniqueness in design feature histories , 1999, SMA '99.

[11]  Kuo-Chin Fan,et al.  Genetic-based search for error-correcting graph isomorphism , 1997, IEEE Trans. Syst. Man Cybern. Part B.

[12]  Carl Ebeling,et al.  SubGemini: Identifying SubCircuits using a Fast Subgraph Isomorphism Algorithm , 1993, 30th ACM/IEEE Design Automation Conference.

[13]  K.L. Kodandapani,et al.  A Wirelist Compare Program for Verifying VLSI Layouts , 1986, IEEE Design & Test of Computers.

[14]  Vincent A. Cicirello.  Intelligent Retrieval of Solid Models , 1999 .

[15]  E. Fantino.  The Analysis of Behavior. , 1971 .

[16]  Dana S. Nau,et al.  Feature-based similarity assessment of solid models , 1997, SMA '97.

[17]  Q. Henry Wu,et al.  Optimization of control parameters in genetic algorithms: a stochastic approach , 1999, Int. J. Syst. Sci..

[18]  James A. Hendler,et al.  The Case for Graph-Structured Representations , 1997, ICCBR.

[19]  E. K. WONG,et al.  Model matching in robot vision by subgraph isomorphism , 1992, Pattern Recognit..

[20]  Jin H. Kim,et al.  Recognizing 3-D objects by forward checking constrained tree search , 1992, Pattern Recognit. Lett..

[21]  John J. Grefenstette,et al.  Optimization of Control Parameters for Genetic Algorithms , 1986, IEEE Transactions on Systems, Man, and Cybernetics.

[22]  Mark F. Bramlette.  Initialization, Mutation and Selection Methods in Genetic Algorithms for Function Optimization , 1991, ICGA.

[23]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[24]  William J. Christmas,et al.  Structural Matching in Computer Vision Using Probabilistic Relaxation , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[25]  Salih O. Duffuaa,et al.  A Linear Programming Approach for the Weighted Graph Matching Problem , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[26]  Walter F. Bischof,et al.  Rulegraphs for graph matching in pattern recognition , 1994, Pattern Recognit..