| 引用本文: | 肖平,余英林.一种基于最大—乘积型合成神经元的模糊联想记忆网络*[J].控制理论与应用,1997,14(6):902~906.[点击复制] |
| XIAO Ping and YU Yinglin.A Kind of Fuzzy Associative Memories Networks Based on Max-Product Composition Units[J].Control Theory and Technology,1997,14(6):902~906.[点击复制] |
|
| 一种基于最大—乘积型合成神经元的模糊联想记忆网络* |
| A Kind of Fuzzy Associative Memories Networks Based on Max-Product Composition Units |
| 摘要点击 1056 全文点击 529 投稿时间：1996-07-22 修订日期：1996-12-23 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 |
| 1997,14(6):902-906 |
| 中文关键词 模糊神经网络 模糊联想记忆 广义模糊解 |
| 英文关键词 neural network fuzzy associative memory connection weight matrix |
| 基金项目 |
|
| 中文摘要 |
| 本文提出了由最大－乘积型合成神经元的模糊神经网络实现双向双向联想记忆的一种学习方法实现双向联想记忆的充要条件，对于自联想记忆网络，自联想权得由广义模糊解确定，模式联想一次就收敛，该网络具有较强的容错性，大量的计算机实验结果表明该学习算法是行之有效的。 |
| 英文摘要 |
| A kind of fuzzy associative memories and its learning algorithm with max-product composition units is presented in this paper. The necessary and sufficient condition to realize bidirectional associative memories is given. The patterns for auto-memories can be convergenced through the connective weight matrix for iteration just once. The connection weight matrix for auto-memories is determined by general fuzzy solution. The computer simulation results have shown the efficiency of the fuzzy network and its learning algorithm. |