基于主成分分析朴素贝叶斯和决策树比较  

The comparison of Naive Bayesian and Decision Tree Based on Principal Component Analysis

作　　者：张林[1] 石洪波[1] ZHANG Lin, SHI Hong-bo（Information Management, Shanxi University of Finance ＆ Economics, Taiyuan 030006, China）

机构地区：[1]山西财经大学信息管理学院,山西太原030006

出　　处：《电脑知识与技术》2010年第1期222-224,共3页Computer Knowledge and Technology

基　　金：国家自然基金项目（60503017）; 山西省自然基金项目（20051046）

摘　　要：朴素贝叶斯和决策树由于其较高的分类性能和简单性得到了广泛的使用,许多学者都在研究如何在分类前对数据进行处理以提升它们的分类性能。该文首先使用主成分分析提取特征数据,然后对处理后的数据上利用朴素贝叶斯和决策树进行分类,并对实验结果进行分析,比较主成分分析对它们分类性能的影响。Naive Bayes and decision tree classifications have been widely used due to their high performance and simplicity, and many scholars are studying how to process the data before classification in order to enhance the performance of their classification.This article first extracted feature data using principal component analysis, and then processed the data on the use of naive Bayes and decision tree classifications, and experimental results were analyzed and compared the impact of the principal component analysis on their classification performance.

关 键 词：朴素贝叶斯分类器 决策树 分类 主成分分析 

分 类 号：TP274[自动化与计算机技术—控制科学与工程;自动化与计算机技术—检测技术与自动化装置]