A New Method for Evolving Artificial Neural Networks Using Genetic Algorithm  

作　　者：Yan Wu Wei Wan 

机构地区：[1]Department of Computer Science and Engineering, Tongji University, Shanghai 200092, China

出　　处：《南昌工程学院学报》2006年第2期79-82,共4页Journal of Nanchang Institute of Technology

基　　金：ProjectSupportedby NSFC (60475019)

摘　　要：In this paper, a new neuroevolution algorithm (NEGA) for simultaneous evolution of both architectures and weights of neural networks is described. A whole new network encoding method is shown. The competing conventions problem is solved absolutely. Heuristic methods are used to constrain the topology mutation probability and the trend of mutation kind choice. Also, the niching method is used to protect the network topologies evolution. The experiment results show the efficiency and rapidity of NEGA forcefully.In this paper, a new neuroevolution algorithm （NEGA） for simultaneous evolution of both architectures and weights of neural networks is described. A whole new network encoding method is shown. The competing conventions problem is solved absolutely. Heuristic methods are used to constrain the topology mutation probability and the trend of mutation kind choice. Also, the niching method is used to protect the network topologies evolution. The experiment results show the efficiency and rapidity of NEGA forcefully.

关 键 词：NEUROEVOLUTION genetic algorithm neural network niching method 

分 类 号：TV[水利工程]