|
|||||
|
|
| 联想Hopfield神经网络样本稳定性的判别方法 | |
| 引用本文： | 赵婷婷,魏小鹏,陈刚.联想Hopfield神经网络样本稳定性的判别方法[J].辽宁工程技术大学学报(自然科学版),2004,23(1):135-137. |
| 作者姓名： | 赵婷婷 魏小鹏 陈刚 |
| 作者单位： | 1. 大连理工大学,机械工程系,辽宁,大连,116023 2. 大连大学,先进设计技术中心,辽宁,大连,116622 3. 辽宁工程技术大学,基础部,辽宁,阜新,123000 |
| 基金项目： | 国家自然科学基金资助(60174037,50275013) |
| 摘 要： | 针对基于Hebb学习的离散Hopfield神经网络，论述了实现联想记忆功能的基本方法及其样本稳定性的基本思想。根据所存在的问题，提出了判别样本稳定性的充要条件，以及判别样本稳定的一个充分条件，使判别样本是否稳定变得简单易行。最后给出了判别样本稳定的矩阵表示形式，并对给出的定理以及推论给予了证明。 |
| 关 键 词： | Hopfield神经网络 样本稳定性 Hebb学习规则 联想记忆 矩阵 |
| 文章编号： | 1008-0562(2004)01-0135-03 |
| 修稿时间： | 2003年4月15日 |
New distinguishing method of prototype stability about associative Hopfield neural networks |
|
| ZHAO Ting-ting,WEI Xiao-peng,CHEN Gang.New distinguishing method of prototype stability about associative Hopfield neural networks[J].Journal of Liaoning Technical University (Natural Science Edition),2004,23(1):135-137. | |
| Authors: | ZHAO Ting-ting WEI Xiao-peng CHEN Gang |
| Institution: | ZHAO Ting-ting1,WEI Xiao-peng2,CHEN Gang3 |
| Abstract: | In view of discrete Hopfield neural networks based on Hebb learning law, the essential method of achieving association and memory function and the essential thinking of prototype stability are discussed. According to the existing problem, the necessary and sufficient conditions of distinguishing prototype stability were put forward. A sufficient condition of distinguishing prototype stability was given. This makes it very easy to distinguish prototype stability. At last, the matrix form of distinguishing prototype stability was given. All theorems and deductions have been proved. |
| Keywords: | hopfield neural networks hebb learning law prototype stability |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |