连续学习混沌神经网络的研究  

Study on Successive Learning Chaotic Neural Network

作　　者：段书凯[1] 刘光远[1] 

机构地区：[1]西南师范大学电子与信息工程系,重庆400715

出　　处：《计算机科学》2004年第4期135-136,139,共3页Computer Science

基　　金：重庆市应用基础基金

摘　　要：近几年混沌神经网络在信息处理,特别是联想记忆中的应用得到了极大重视。本文提出了一个改进的连续学习混沌神经网络(MSLCNN)模型,它具有两个重要特征:(1)根据不同的输入,神经网络做出不同的响应,可从已知模式来识别未知模式;(2)可连续学习未知模式。计算机仿真表明我们的模型具有应用潜力。The applications of chaotic neural network in the field of message treatment especially in associative memories have drawn a lot of attention in recent years. In this paper, we propose a modified successive learning chaotic neural network (MSLCNN)model. It has two important features: (l)It can distinguish unknown patterns from the known patterns according to different response of the neural network when different input applied. C2)It can learn unknown pattern successively. When a stored pattern is given to the network, the network searches around the input pattern. However, while an unknown pattern is given, a chaotic itinerancy appears. The MSLCNN makes use of these features to distinguish unknown patterns from the known patterns and learn the unknown patterns successively. A series of computer simulations are done to demonstrate the potentiality of the proposed model.

关 键 词：连续学习 混沌 神经网络 GCM模型 计算机仿真 信息处理 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391.9[自动化与计算机技术—控制科学与工程]