Abstract:
Two evolutionary learning algorithms for feed forward neural netWork are presented in this paper,one is GA,the other is EP.According to this conclusion, two evolutionary learning algorithms are configured, and get far superior performance over traditional BPalgorithm.
摘要: 提出两种用于前向神经网络的进化学习算法,一种基于遗传算法,另一种基于进化规则。通过对XOR问题和IRIS模式分类问题的学习,证明它们远高于传统BP算法的性能。
关键词:
种经网络,
进化计算,
遗传算法,
进化规划
CLC Number:
Wang Lei, Qi Feihu. Application of Evolutionary Computation in Neural Network[J]. , 1999, 25(11): 41-43.
王磊,戚飞虎. 进化计算在神经网络学习中的应用[J]. , 1999, 25(11): 41-43.