基于关联度分析的遗传算法研究及其应用

Research and Application of Correlation-based Genetic Algorithm

简要分析了关联规则数据挖掘技术的研究现状,给出了关联规则问题描述,引出了基于关联度的适应度函数,最终建立了一个基于关联度分析的遗传算法。并应用于银行卡分析系统,建立特征提取的具体实例即客户信用度预测模型。实例结果表明具有较好的分类预测能力。

This paper analyses research status in quo of correlation-based data mining technology, gives the definition of correlation rule and advances the correlation-based fitness function. Then this paper makes a new genetic algorithm which is based on correlation rule, and apples it to the credit card analysis system. An example of feature selection is performed, the result shows that this algorithm has good predictive ability.