形态学联想记忆网络的研究进展  

Research summary on morphological associative memory networks

作　　者：冯乃勤[1] 李素娟[1] 敖连辉[1] 王双喜[1] 王岁花[1] 

机构地区：[1]河南师范大学计算机与信息技术学院,河南新乡453007

出　　处：《计算机工程与设计》2010年第21期4665-4669,4673,共6页Computer Engineering and Design

基　　金：河南省自然科学基金项目(102300410099);河南师范大学博士基金项目(521662)

摘　　要：形态学联想记忆网络具有无限存储能力、一步回忆记忆、良好地抵抗腐蚀噪声或者膨胀噪声的噪声容限等许多优点。从形态学联想记忆的概念、基本原理、发展脉络、研究新成果、发展趋势和研究方向等多个方面综述了形态学联想记忆网络的研究进展。对形态学联想记忆方面的研究带来了一定的参考价值。The morphological associative memory networks have many attractive advantages, for example, unlimited storage capacity; one-shot recall speed and good noise-tolerance to single erosive or dilative noise. In the past 20 years, morphological associative memory networks as an important branch of morphological neural networks have undergone considerable development, and timely review is necessary. The advances of the research on morphological associative memory networks is summarized from the following aspects in- cluding concepts, basic principles, developmental skeleton, new research findings, development trend and research direction etc. It＇ s expected to be of great benefit to the research on morphological associative memory networks.

关 键 词：联想记忆 形态学联想记忆 神经网络 异联想 模式识别 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]