一种基于正区域及其分类纯度的决策树算法  

An Algorithm for Decision Tree Based on Positive Region and Its Classification Ability

作　　者：蒋慧平[1,2] 费耀平[3] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410012 [2]湖南广播电视大学,湖南长沙410004 [3]中南大学现代教育技术中心,湖南长沙410012

出　　处：《湖南广播电视大学学报》2011年第3期57-59,共3页Journal of Hunan Radio and Television University

摘　　要：本文分析了基于正区域的决策树生成算法的不足,针对这些不足,提出了基于正区域及其分类纯度的决策树算法。该方法计算简单,易于理解,并用实例说明了该方法的优越性。In this paper, disadvantages of algorithm for constructing decision tree based on positive region are analyzed. Aiming at these disadvantages, a new algorithm for decision tree based on positive region and its classification ability is proposed. The method is simple and easy- to- understand. The example shows that the method is effective.

关 键 词：粗糙集 决策树 正区域 分类纯度 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]