Several symmetry properties of discrete Hopfield neural networks

Proceedings. International Conference on Machine Learning and Cybernetics Pub Date : 2002-11-04 DOI:10.1109/ICMLC.2002.1167431

Jiyang Dong, Jun-ying Zhang

引用次数: 2

Abstract

Symmetry is powerful tool to reduce the freedom of a problem. Discrete Hopfield neural networks with Hebbian learning are studied by the method of group theory in this paper, and several symmetry properties of the network being an auto-associator are given and proved.

查看原文本刊更多论文

离散Hopfield神经网络的几个对称性

对称是减少问题自由度的有力工具。本文用群论的方法研究了具有Hebbian学习的离散Hopfield神经网络，给出并证明了该网络为自关联子的几个对称性质。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Proceedings. International Conference on Machine Learning and Cybernetics

自引率

0.00%

发文量