神经元模式分类器学习的进化计算算法

本文提出了一种用于神经元模式分类器学习的进化计算算法。该算法综合了非确定有限自动机和次群体的动态数据结构，可有效地完成神经网络模式分类器的结构学习，以获得最优的求解结果。该算法的有效性已由计算机仿真实验所证实，可被认为是一种很有发展前途的模式分类系统的机器学习算法。

AN EVOLUTIONARY AIGORIIHM FOR LEARNING OF A NEURAL PATIERN CLASSIFIER

This paper proposes an evolutionary algorithm for learning of a neural pattern classifier.This algorithm integrates non-determinate finite automata and dynamic data structure of sub-population. It can perform structure learning for a neural network pattern classifier efficiently to obtain the optimal result of problem solving. Its effectiveness has been testified by computer simulating experiment and it can be regarded as a promising machine learning for the pattern classification system.