摘要
客户分类是智能CR M系统的一个重要功能。该文提出使用前馈型神经网络构作一个CRM客户分类模型的思想,并用免疫遗传算法对其进行优化。在染色体设计上提出了三层结构的染色体,解决单层结构染色体中,短基因组实际变异机率过小的问题,促使子代种群有更好的“生物多样性”。最后通过实验证明模型有较好的客户类别识别率及优点。
The customer classifier is an important part of an Intelligent CRM system.In this article,we propose a method of designing a customer classifier based on feedback neural network,and optimize it with immune genetic algorithm.A chromosome with three layers is proposed to solve the problem that a shorter gene group has fewer opportunities to mutate.At last,our experiment shows that this model has a better rate on customer classifying and more excellences.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第14期182-184,共3页
Computer Engineering and Applications
基金
广西教育厅桂教科研2003(编号:22号)
