Zheng Pengsheng, Zhang Jianxiong, Tang Wansheng
Institute of Systems Engineering, Tianjin University, Tianjin 300072, China.
IEEE Trans Neural Netw. 2011 Mar;22(3):347-55. doi: 10.1109/TNN.2010.2099239. Epub 2010 Dec 23.
In this paper, a method for the design of Hopfield networks, bidirectional and multidirectional associative memories with asymmetric connections, is proposed. The given patterns can be assigned as locally asymptotically stable equilibria of the network by training a single-layer feedforward network. It is shown that the robustness in respect to acceptable noise in the input of the constructed networks is enhanced as the memory dimension increases and weakened as the number of the stored patterns grows. More important is that the remembered patterns are not necessarily of binary forms. Neural associative memories for storing gray-level images are constructed based on the proposed method. Numerical simulations show that the proposed method is efficient for the design of Hopfield-type recurrent neural networks.
本文提出了一种具有非对称连接的双向和多向联想记忆的霍普菲尔德网络设计方法。通过训练单层前馈网络，可以将给定模式指定为网络的局部渐近稳定平衡点。结果表明，随着记忆维度的增加，所构建网络输入中可接受噪声的鲁棒性增强，而随着存储模式数量的增加，鲁棒性减弱。更重要的是，记忆模式不一定是二进制形式。基于该方法构建了用于存储灰度图像的神经联想记忆。数值模拟表明，该方法对于霍普菲尔德型递归神经网络的设计是有效的。
