主动联想记忆神经网络模型

在信息编码能提高联想记忆的存贮能力和脑内存在主动活动机制的启发下，提出一个主动联想记忆模型。模型包括两个神经网络，其一为输入和输出网络，另一个为在学习时期能自主产生兴奋模式的主动网络。两个网络的神经元之间有突触联系。由于自主产生的兴奋模式与输入无关，并可能接近于相互正交，因此，本模型有较高的存贮能力。初步分析和计算机仿真证明：本模型确有比通常联想记忆模型高的存贮能力，特别是在输入模式间有高度相关情况下、最后，对提出的模型与双向自联想记忆和光学全息存贮机制的关系作了讨论。

AN ACTIVE ASSOCIATIVE MEMORY NEURAL NETWORK MODEL

Inspired by the facts that the information coding can improve the capacity of associative moemory and that brain has active mechanism,an active mechanism,an active odetive memory model is propose.The model consists of two neural networks. One is a network of input-output and the other can produce endogenous patterns in loarning phase.There are synaptic connections between the neurons of two networks.Hewever,the endogenous patterns are indepenent from input patterns and be approxhately orthogonal each other,so that the model has higher performance. Preliminary analysis and computer simulatin show that the model possesses muchhigher capacity then that of conventional associative memory models, specially in the case which input patterns are highly correlative.Final,the relationship between the model and bidirectional associative memory and hologram are discussed.