Affiliation:
1. Wuhan University
2. Jingdezhen Ceramic Institute
3. South China University of Technology
Abstract
There are many similarities between the intelligence algorithms and the Statistical Physics and Thermodynamics in many aspects, such as the research object, tasks and methods. Firstly, this paper presents the hybrid intelligence algorithms improved by Thermodynamics. Then the theory analysis of intelligence algorithms by Thermodynamics is presented. Finally, a new research direction, that is the novel intelligence algorithm based on statistical physics and thermodynamics, is proposed for the future.
Publisher
Trans Tech Publications, Ltd.
Reference24 articles.
1. Dayou Wu. Thermodynamics, kinetic theory of gases and statistical mechanics [M]. Beijing: Science Press, (1983).
2. S. Kirkpatrick, C. D. Gelatt, Jr., M. P. Vecchi. Optimization by Simulated Annealing [J]. Science, 1983, 220(4598): 671-680.
3. Mori N, Yoshida J, Tamaki H, Kita H, et al. A thermodynamical selection rule for the genetic algorithm [C]. IEEE Conf. on Evolutionary Computation. New Jersey: IEEE Press, 1995. 188-192.
4. W. Ying, Y. Li, SHEU Phillip C-Y et al. Geometric Thermodynamical Selection for Evolutionary Multi-Objective Optimization [J]. Chinese Journal of Computers. 2010 33(4) 755-767.
5. K. Li, Y. Li, L. Kang, et al. A Multi-objective Evolutionary Algorithm Based on Transportation Theory [J]. Chinese Journal of Computers. 2007 30(5) 796-805.