案例属性权重的群基数效用优化方法

摘要

案例特征属性权重的选取直接影响到案例检索的精度．针对属性权重的优化选择问题，提出一种基于群决策思想的案例属性权重优化方法．首先，利用遗传算法从已有的案例数据库中得出多组初始特征属性权重；然后，借鉴群决策思想中的群基数效用法对多组权重进行优化选择，并在推理过程中进行权重的自适应调整，以此确定合理的特征属性权重．仿真结果表明，该方法应用于案例推理过程能充分挖掘出多组权重的潜在知识，从而得出精度更高的检索结果．%The selection of the case feature attribute weights directly affects the case retrieval precision. According to the existing problem of optimizing attribute weights, a new method based on group decision- making thought for optimizing the case feature attribute weights is proposed in this paper. Multiple sets of initial feature attribute weights are first obtained by genetic algorithm. The multiple sets of weights by group cardinal utility method are then optimized, and the weights can be adaptively adjusted during the reasoning process to ascertain reasonable feature attribute weights. Simulation results show that the proposed approach can fully excavates the potential knowledge that exists in multiple sets of attribute weights and improves the retrieval precision of the case-based reasoning system.