基于属性聚类的贝叶斯分类算法

Bayes Classification Algorithm Based on Attribute Clustering

在数据挖掘的过程中,由于贝叶斯分类算法要求条件属性必须独立,因此具有一定的局限性.考虑该缺陷,提出用数据挖掘中另一种常用聚类算法对原始数据的条件属性进行数据预处理后,再使用贝叶斯算法.实验表明该算法提高了分类的准确率.

In data mining process, Bayes classification algorithm requires condition attributes be independent, so it has some limits. According to the limitation, clustering algorithm is used to preprocess the condition attribute before Bayes classification algorithm is proposed. Experiment shows that algorithm improves the accuracy of classification.