摘要
本文从函数映射关系来分析似然推理的三段论形式，结合似然推理的公理化约束．证明了该映射关系为连续形式．因此提出用带有判决层的前馈三层神经网来实现模糊似然推理．用大前提作依据，产生近似规则来进行训练，并给出例子证明该算法的有效性．
In this paper,the functional mapping was used for analyzing the likelihood inferences of syllogism, The continuous type of mapping relationship was identified by combining with empirical comstraint of likelihood inferences. A three-layer forward feedback neural network with a decision layer layer was used for realizing a fuzzy likelihood decision. The generated approximate rules based the major premise was used as training samples. Finally the effects of this proposed algorithm was verified by simulation results.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
1996年第2期182-190,共9页
Control Theory & Applications
基金
攀登计划(模糊神经网络模型)资助
关键词
神经网络
模糊系统
似然推理
知识处理
neural network
fuzzy logic: likelihood inference: knowledge Processing
