论文部分内容阅读
形态学联想记忆是一类新颖的人工神经网络.尽管它有许多优点,然而也有一个致命的弱点,即异联想记忆效果较差.通过对形态学联想记忆框架的分析,发展了一种新的形态学联想记忆方法——加减形态学联想记忆(MAMFATS).理论分析和实验证实,该方法对于异联想形态学记忆效果显著,在一定程度上可弥补原形态学联想记忆方法的不足.新旧两类方法之间的关系是互补,而非排斥.两种方法结合起来,可以更好应对异联想形态学记忆不完全回忆记忆的困难.MAMFATS方法充实了形态学联想记忆理论,也增大了异联想形态学记忆的实用性.
Morphology Associative memory is a kind of novel artificial neural network, and although it has many advantages, it has a fatal weakness, ie, the effect of memorizing memory is poor.Analyzing the morphological associative memory framework, a new Morphological Associative Memorization - Modified Morphological Associative Memorization (MAMFATS). The theoretical analysis and experimental results show that this method has significant effect on the morphological memory of allogeneic associations and to a certain extent can make up for the deficiencies of the original morphological associative memory methods. The relationship between the two methods is complementary rather than exclusive.The combination of the two methods can better cope with the difficulty of incompletely remembering the memory of the X-linked morphological memory.MAMFATS method enriches the morphological associative memory theory and increases The Practicality of Different Legendary Morphological Memory.