摘要
提出了不确定决策的双向联想记忆方法。这种方法通过关联矩阵将归纳好的一些决策规则存储到记忆矩阵中。当有新的状态信息输入时，通过对信息流的前向处理和后向处理，使它收敛到一个稳定的规则，从而给出合理的决策方案。应用实例说明了这种方法的有效性。
In this paper, a bidirectional associative memory based approach is proposed, which can be used to deal with uncertain decision making. In this approach, come decision rules are stored in an associative matrix, by processing the new inputted state information forward and backward, a reasonable decision alternative can be obtained. A numerical example is discussed to demonstrate the implementation of the proposed approach.
出处
《南京邮电学院学报》
1997年第4期162-164,共3页
Journal of Nanjing University of Posts and Telecommunications(Natural Science)
基金
邮电部重点科研基金
关键词
人工智能
决策分析
神经网络
双向联想记忆方法
Artificial intelligence, Decision analysis, Neural networks, Bidirectional associative memory approach
