Human SCINT Seminar (9)
Event Date: 2005-07-27 13:00
| Date: 2005.7.27 (Wed) 13:00-14:15
Place: General Research Building, Room 663
Speaker: Kohsuke Yanai
Title: Emergence of intelligence from evolutionary computing
Keywords: evolutionary computing, program evolution, emergence, artificial life, complex system biology
Affiliation: Department of Frontier Informatics, Graduate School of Frontier Sciences
Position: Graduate Student
Adviser: Hitoshi Iba, Iba laboratory
Disciplines: Informatics, artificial intelligence, evolutionary computing
Societies and Conferences: The Japanese Society for Artificial Intelligence, Information processing society of Japan, ACM SIGEVO , Genetic and Evolutionary Computing Conference, Congress on Evolutionary Computation, Asian-Pacific Workshop on Genetic Programming
Bibliography: Kohsuke Yanai, Emergence of intelligence from evolutionary computing, Human Science Integration Seminar Abstracts, No. 9, pp. 1, 2005.
(Please use this bibliography when you cite this abstract.)
Abstract:
Can we gather any kind of intelligence or knowledge through simulations of evolution? Evolutionary computation is an approach based on natural evolution for problem-solving and optimization. It uses the Darwinian principle of natural selection and analogs of various naturally occurring operations, including crossover, mutation, gene duplication, gene deletion.
Evolutionary methods for automatically creating a computer program are one of the main topics in the field of evolutionary computation. There are now 37 instances where these methods have produced human-competitive results: 22 instances where they have created the functionality of a previously patented invention, and 2 instances where they have created a patentable new invention.
In this presentation, we overview evolutionary computation and its applications to program generation. Then we describe evolutionary algorithms that use neither crossover nor mutation but uses probability distributions, and review recent studies concerning an approach based on a probabilistic model. Finally we discuss the mechanism of evolutionary computation from a viewpoint of "Complex Systems."
References:
[Beyer 05] Beyer, H.-G. ed.: Genetic and Evolutionary Computation ConferenceACM SIGEVO (2005)
[Dawkins 76] Dawkins, R.: The Selfish Gene, Oxford University Press, 1 edition (1976)
[Iba 99] Iba, H. and Sasaki, T.: Using Genetic Programming to Predict Financial Data Series, in IEEE Conference on Evolutionary Computation (CEC99), IEEE Press (1999)
[Ikegami 95] Ikegami, T. and Hashimoto, T.: Coevolution of Machines and Tapes, in Advances in Artificial Life, 234–254 (1995)
[Kimura 05] Kimura, S. and Matsumura, K.: Genetic Algorithms using Low-Discrepancy Sequences, in Genetic and Evolutionary Computation Conference, pp.1341–1346 (2005)
[Koza 92] Koza, J. R.: Genetic Programming: On the Programming of Computers by Means of Natural Selection, MIT Press (1992)
[Koza 03] Koza, J. R., Keane, M. A., Streeter, M. J., Mydlowec, W., Yu, J., and Lanza, G.: Genetic Programming IV: Routine Human-Competitive Machine Intelligence, Springer, 1 edition (2003)
[Larranaga 02] Larranaga, P. and Lozano, J. A.: Estimation of Distribution Algorithms, Kluwer Academic Publishers (2002)
[Sastry 03] Sastry, K. and Goldberg, D. E.: Probabilistic model building and competent genetic
programming, in Genetic Programming Theory and Practise, pp. 205–220, Kluwer (2003)
[Shan 04] Shan, Y., McKay, R., Baxer, R., Abbass, H., Essam, D., and Nguyen, H.: Grammar Model-based Program Evolution, in Proceedings of the Congress on Evolutionary Computation: CEC-2004, pp. 478–485 (2004)
[Yanai 01] Yanai, K. and Iba, H.: Multi-agent Robot Learning by Means of Genetic Programming : Solving an Escape Problem, in Evolvable Systems : From Biology to Hardware, Springer-Verlag (2001)
[Yanai 03] Yanai, K. and Iba, H.: Estimation of Distribution Programming based on Bayesian Network,
in Proceedings of Congress on Evolutionary Computation: CEC-2003, pp. 1618–1625 (2003)
[Yanai 04] Yanai, K. and Iba, H.: Program Evolution by Integrating EDP and GP, in Proceedings of Genetic and Evolutionary Computation Conference GECCO-2004 (2004)
[Yanai 05] Yanai, K. and Iba, H.: Probabilistic Distributio Model of EDA-based GP, Genetic and Evolutionary Computation Conference (2005) |
|