通过采用混合CBR方法研究了以客户分类为代表的商务分类问题。该方法在检索前引入遗传算法(GA)优化案例特征项权重,避免了传统k-NN法的不足,并以某零售业销售数据库为例进行了实验应用。
This paper utilizes the hybrid CBR approach to business classification represented by customer classification. It overcomes the disadvantage of conventional k - NN approach by the way of introducing Genetic Algorithm (GA) into the optimization of case feature weights before retrieval. Further, it experiments with a retailing sales database as example, whose result supports GA -CBR application.