首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于免疫算法的前向神经网络学习方法
引用本文:宫新保,臧小刚,周希朗.基于免疫算法的前向神经网络学习方法[J].系统工程与电子技术,2004,26(12):1927-1929.
作者姓名:宫新保  臧小刚  周希朗
作者单位:上海交通大学电子工程系,上海,200030
摘    要:提出了一种采用免疫算法训练多层前向神经网络的方法。该方法利用免疫算法训练前向神经网络,能够使网络优化过程趋于全局最优。利用基于遗传策略的聚类机制确定前向神经网络的初始权值,增加了网络训练算法收敛于全局最优的概率。将这种神经网络用于雷达模拟调制信号的调制方式识别的仿真结果表明,采用该算法设计的前向神经网络达到了较高的性能。

关 键 词:前向神经网络  免疫算法  人工免疫机制  遗传聚类算法
文章编号:1001-506X(2004)12-1927-03
修稿时间:2003年8月21日

Learning strategy of feedforward neural network based on immune algorithm
GONG Xin-bao,ZANG Xiao-gang,ZHOU Xi-lang.Learning strategy of feedforward neural network based on immune algorithm[J].System Engineering and Electronics,2004,26(12):1927-1929.
Authors:GONG Xin-bao  ZANG Xiao-gang  ZHOU Xi-lang
Abstract:A method to design the multi-layer feedforward neural network based on immune algorithm is proposed. In this method, immune algorithm is used to design the network that makes the training process tending to global optima. A genetic clustering algorithm is used to determine the initial weight vectors, therefore the probability of training algorithm to converge to global optima is improved. The applications of the neural network in the modulation-style recognition of analog modulated radar signals demonstrates a good performance of the network.
Keywords:feedforward neural network  immune algorithm  artificial immunology  genetic clustering algorithm
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号