摘要
分类算法一直以来都是数据挖掘领域的研究重点,朴素贝叶斯分类算法是众多优秀分类算法之一,但由于其条件属性必需独立,使得该算法也存在着一定的局限性。为了从另外一种角度来改进该算法,提高分类性能,提出了一种基于K-近邻法的局部加权朴素贝叶斯分类算法。使用K-近邻法对属性加权,找到最合适的加权值,运用加权后的朴素贝叶斯分类算法去分类,实验表明该算法提高了分类的可靠性与准确率。
Classification algorithm has been the focus of research in the field of data mining,the Naive Bayes classification algorithm is one of the good classification algorithms.Because its condition attributes shall be independent however,there are some limitations in the algorithm.In order to improve the classification performance of the algorithm from another side,the locally weighted Naive Bayes classification algorithm based on K-nearest neighbour has been proposed in this paper.K-nearest neighbour method is used to weight the attributes to find the appropriate weights,the weighted Naive Bayes classification algorithm is then used for classification.Experiment shows that the algorithm improves the reliability and accuracy of the classification.
出处
《计算机应用与软件》
CSCD
2011年第9期267-268,291,共3页
Computer Applications and Software
关键词
朴素贝叶斯
K-近邻法
局部加权
分类
Naive Bayes K-nearest neighbour Locally weighted Classification
