由天津大学系统工程研究所研制的国家自然科学基金项目“具有认知功能的人脑模型(神经网络模型)的研究,经中科院计算所等单位鉴定达到了国际同类研究的先进水平。 该项目的研究内容包括:(1)具有人脑功能的联想记忆研究;(2)基于神经网络研究的一种思维的潜运算模型探索;(3)对BSB模型的概念识别能力的推广研究。 本项研究的主要成果是提出了三种联想记忆模型:(1)具有最优联接矩阵的前向异联想模型。该模型利用输入层神经元间的相互抑制作用,只用两层神经元即可实现最优联想映射,较迄今已提出的三层网络 The project of National Natural Science Foundation of China (NSFC) developed by Institute of Systems Engineering of Tianjin University "The research of human brain model with cognitive function (neural network model) has reached the advanced level of international researches by the Institute of Computing, Chinese Academy of Sciences The research contents include: (1) associative memory research with human brain function; (2) exploration of a thinking latent computing model based on neural network research; (3) generalization of the conceptual recognition ability of BSB model. The main achievement is to propose three kinds of associative memory models: (1) Forward heteroconversation model with the optimal connection matrix, which takes advantage of the mutual inhibition of the input layer neurons and only uses two layers of neurons Excellent association mapping, so far has been proposed three-tier network