ABSTRACT
This paper proposes a novel technique for a program evolution based on probabilistic models. In the proposed method, two probabilistic distribution models with probabilistic dependencies between variables are used together. We empirically comfirm that our proposed method has higher search performance. Thereafter, we discuss the effectiveness of its distribution models.
- R. P. Salustowicz and J. Schmidhuber. Probabilistic incremental program evolution: Stochastic search through program space. In M. van Someren and G. Widmer, editors, Machine Learning: ECML-97, volume 1224, pages 213--220. Springer-Verlag, 1997. Google ScholarDigital Library
- K. Sastry and D. E. Goldberg. Probabilistic model building and competent genetic programming. In Genetic Programming Theory and Practise, pages 205--220. Kluwer, 2003.Google Scholar
- Y. Shan, R. McKay, R. Baxer, H. Abbass, D. Essam, and H. Nguyen. Grammar model-based program evolution. In Proceedings of the Congress on Evolutionary Computation: CEC-2004, pages 478--485, 2004.Google ScholarCross Ref
- K. Yanai and H. Iba. Estimation of distribution programming based on Bayesian network.InProceedings of Congress on Evolutionary Computation: CEC-2003, pages 1618--1625, 2003.Google Scholar
Index Terms
- Probabilistic distribution models for EDA-based GP
Recommendations
Probabilistic model-building genetic algorithms
GECCO '08: Proceedings of the 10th annual conference companion on Genetic and evolutionary computationProbabilistic model-building algorithms (PMBGAs) replace traditional variation of genetic and evolutionary algorithms by (1) building a probabilistic model of promising solutions and (2) sampling the built model to generate new candidate solutions. ...
Probabilistic model-building genetic algorithms
GECCO '10: Proceedings of the 12th annual conference companion on Genetic and evolutionary computationProbabilistic model-building algorithms (PMBGAs) replace traditional variation of genetic and evolutionary algorithms by (1) building a probabilistic model of promising solutions and (2) sampling the built model to generate new candidate solutions. ...
Probabilistic model-building genetic algorithms
GECCO '09: Proceedings of the 11th Annual Conference Companion on Genetic and Evolutionary Computation Conference: Late Breaking PapersProbabilistic model-building algorithms (PMBGAs) replace traditional variation of genetic and evolutionary algorithms by (1) building a probabilistic model of promising solutions and (2) sampling the built model to generate new candidate solutions. ...
Comments