摘要
本文首先阐述了神经网络中的一般更新规则，着重讨论了两类模型的更新规则，在此基础之上提出新的单元更新规则，通过利用ＴｕｒｂｏＣ语言实现由此得到的二个模型，并实际解决了皇后问题，ＴＳＰ等例子，解题效果是良好的。
This paper describes general update rule in Neural Networks at first, and emphatically discusses two update rules. Based upon these, the paper puts forward new update rules of unit. Therefore we have got two new models. Through realizing the new models,we have found that the effect of resolving problems with new models is rather good.
出处
《小型微型计算机系统》
CSCD
北大核心
1994年第1期46-50,共5页
Journal of Chinese Computer Systems
基金
863计划的部分资助
关键词
神经网络
更新规则
约束满足模型
Neural networks , Parallel distributed processing, Constraint satisfaction, Update rule
