Presentation 2004-06-21
A new selection method in genetic algorithm(2) : Experiments on De Jong's functions
Satoshi SAWATANI, Tetsuaki HIRAHAYA, Yasuko MUNEHISA, Tomo MUNEHISA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a genetic algorithm which uses a new selection method. This method assumes that a operation of genetic algorithm is a Malcov chain, and a stationary distribution obtained becomes a Boltzmann distribution. This convergence to the Boltzmarm distribution guarantrees the optimal solution. Since we cannot keep all states, we have to drop some states. For this purpose we use a on-off probability function to select states., where this function gives us a probability that we drop or leave a state. Thereby, we can obtain the Boltzmann distribution only holding a few states. We verify a validity of our genetic algorithm applying it to De Jong's functions.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) GeneticAlgorithm / DeJong's Functions / Optimization / Markov chain
Paper # AI2004-10
Date of Issue

Conference Information
Committee AI
Conference Date 2004/6/14(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Artificial Intelligence and Knowledge-Based Processing (AI)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A new selection method in genetic algorithm(2) : Experiments on De Jong's functions
Sub Title (in English)
Keyword(1) GeneticAlgorithm
Keyword(2) DeJong's Functions
Keyword(3) Optimization
Keyword(4) Markov chain
1st Author's Name Satoshi SAWATANI
1st Author's Affiliation Faculty of Engineering, Yamanashi University()
2nd Author's Name Tetsuaki HIRAHAYA
2nd Author's Affiliation Faculty of Engineering, Yamanashi University
3rd Author's Name Yasuko MUNEHISA
3rd Author's Affiliation Faculty of Engineering, Yamanashi University
4th Author's Name Tomo MUNEHISA
4th Author's Affiliation Faculty of Engineering, Yamanashi University
Date 2004-06-21
Paper # AI2004-10
Volume (vol) vol.104
Number (no) 133
Page pp.pp.-
#Pages 6
Date of Issue