全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

进化策略求解Hopfield神经网络

DOI: 10.11834/jig.20000713

Keywords: 模式识别,进化策略,Hopfield神经网络,优化

Full-Text   Cite this paper   Add to My Lib

Abstract:

提出了一种进化策略求解Hopfield神经网络的方法.该进化策略分三个阶段,即第一阶段只在较小区间上求出局部优化解;然后,在此基础上,由第二阶段求出较大区间上的局部优化解;最后由第三阶段求出全局优化解.同时采用Hopfield神经网络动态方程指导第一阶段的局部进化策略的进化方向,因而大大加快了优化搜索速度.在分阶段的进化策略中,其第一阶段只需搜索较小区间,第二和第三阶段的搜索则建立在其前一阶段的基础上,因此可以采用较小的遗传群体规模,从而明显地提高了求解Hopfield神经网络的速度,减少了计算内存的需求.

References

[1]  David B F. An introduction to simulated evolutionary optimization. 1EEE Transactions on Neural Networks,1994,(5):3-14.
[2]  Lai W K,Coghill G G, Genetic breeding of control parameters for the Hopfield/Tank neural net. In: Proc. Of International Joint Conference on Neural Networks ,Baltimare,USA. 1992,618-623.
[3]  Hiroyuki S, Atsushi 1, Shunji K, Tsuneo T. A solution of eombinatorial optimization problem by uniting genetic algorithms with Hoplield\\'s model. In: Proc, of 1994 IEEE International Conference on Neural Networks,Orlando,USA,1994,4704-4709.
[4]  Yao X.A reviews of evolutionary artificial meural networks.International Journal of Intelligent Systems, 1993, (8): 539-567.
[5]  Hopfield J J, Tank D B. "Neural" computation of decisions in optimization problems. Biol. Cybern. 1985,(52):141-152.
[6]  Toshio F,Tadashi K,Takanori S. Heuristic leaning by genetic algorithm for recurrent neural network. In: Proc, of the 32nd 1EEE Conference on Decision and Control,San Antonio,USA,1993,71-77.
[7]  Chao C H,Dhawan A P, Edge detection using a Hopfield neural network, Optical Engineering, 1994, (33): 3739-3747.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133