摘要
信息论是数据挖掘技术的重要指导理论之一,是决策树算法实现的理论依据。决策树算法是一种逼近离散值目标函数的方法,其实质是在实例学习的基础上,得到分类规则。本文简要介绍信息论的基本原理,重点阐述基于信息论的决策树算法,分析了它们目前主要的代表理论以及存在的问题。
The information theory is one of the basic theories of Data Mining, and also is the theoretical foundation of the Decision Tree Algorithm. Decision Tree Algorithm is a method to approache the discrete - valued objective function. The essential of the method is to obtain a classification rule on the basis of example- based learning.
出处
《自动化技术与应用》
2006年第1期4-7,共4页
Techniques of Automation and Applications
关键词
数据挖掘
信息论
决策树
信息熵
Data mining
Information theoretic
Decision tree
Entropy
