摘要
信用为客户选择供应商提供了一个衡量标准.指标体系和评价方法的选择是信用评价的关键.在建立的评价指标基础上,设计的遗传聚类算法通过对案例的强化学习和聚类分析,达到了对供应商信用分类的目的,为信用评价提供了科学的方法.实例验证了该算法的有效性.
Credit provides customers a standard for choosing suppliers. The index system and method are the keys of credit evaluating. On the base of creating the index of evaluation, this paper designs the algorithm of genetic clustering, attains the aim of the credit classification of suppliers by reinforcement learning from cases and clustering analysis, and provides a scientific method for the evaluation of credit. An evaluation instance shows its validity.
出处
《三峡大学学报(自然科学版)》
CAS
2007年第3期242-244,共3页
Journal of China Three Gorges University:Natural Sciences
基金
湖北省教育厅科学技术研究项目(2004D020)
关键词
信用评价
强化学习
聚类分析
遗传算法
evaluation of credit
reinforcement learning
clustering analysis
genetic algorithm