摘要
C C Wang等利用 Kosko的双向联想记忆模型 (Bidirectional associative memory,BAM) ,构造了由多个 BAM构成的多重 BAM(Multi- BAM)决策模型 ,使之可以应用于多证据推理 ,获得了 Multi- BAM的决策性能。作者在此基础上 ,通过对各 BAM引入不同的权值来模拟各专家不同的权威度 ,构建了相应的多重加权 BAM(Multi- WBAM)模型 ,证明了该模型在同、异步方式下的的稳定性 ,并获得其决策性能。
C C Wang and coworkers built a decision making model consisting of multiple bidirectional associative memory (Multi BAM) through BAM with equal priviledge, applied it to the decision making multiple experts, and obtained its decision making performance. In this paper, endowing different privileges to each BAM or expert, a multiple weighted decision making model consisting of BAM (Multi WBAM)is constructed and investigated. Firstly its stability in synchronous and asynchronous updating modes is proved and then its decision making performance and majority factor for different privileges of each expert are obtained. Finally, the results of the given examples and the computer si mulations are coincident with the intuitive human reasoning.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2000年第6期625-630,共6页
Journal of Nanjing University of Aeronautics & Astronautics
基金
国家自然科学基金!(编号 :6970 1 0 0 4 )
南京大学计算机软件新技术国家重点试验室基金资助项目
关键词
神经网络
决策系统
加权
双向联想记忆
neural network
decision making system
weighted
multiple evidence reasoning
bidirectional associative memory
