| 基于模式分量最优稳定性准则的联想记忆学习算法 |
| |
| 引用本文： | 梁学斌,吴立德.基于模式分量最优稳定性准则的联想记忆学习算法[J].电子学报,1996,24(5):30-33. |
| |
| 作者姓名： | 梁学斌 吴立德 |
| |
| 作者单位： | 复旦大学计算机科学系 |
| |
| 基金项目： | 国家攀登计划资助,国家863一512主题资助 |
| |
| 摘 要： | 提出了设计联想记忆神经网络的模式分量最优稳定性准则，即使得每个模式分量对于模式畸变都具有最大的联想容错性。
|
| 关 键 词： | 联想记忆 最优稳定性 学习算法 神经网络 |
A Learning Algorithm of Associative Memory with Optimal Stability of Pattern Components |
| |
| Liang Xuebin, Wu Lide.A Learning Algorithm of Associative Memory with Optimal Stability of Pattern Components[J].Acta Electronica Sinica,1996,24(5):30-33. |
| |
| Authors: | Liang Xuebin Wu Lide |
| |
| Abstract: | The criterion which requires each pattern component to possess optimal stability is proposed. Based on the criterion,a constrained perceptron optimization algorithm is developed,which has the following two characteristics: (1) It can store any given training patterns set; (2)Each pattern component possesses optimal stability. Computer simulation results confirm the ad vantages of our optimization learning algorithm over existing ones. |
| |
| Keywords: | Associative memory Total storage abibity Optimal stability Optimal algorithm |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
