模糊决策树算法与清晰决策树算法的比较研究

ID3算法是一种典型的决策树归纳算法,这种算法在假定示例的属性值和分类值是确定的前提下,使用信息熵作为启发式建立一棵清晰的决策树。针对现实世界中存在的不确定性,人们提出了另一种决策树归纳算法,即模糊决策树算法,它是清晰决策树算法的一种推广。这两种算法在实际应用中各有自己的优劣之处,针对一个具体问题的知识获取过程,选取哪一种算法目前还没有一个较明确的依据。该文从5个方面对这两种算法进行了详细的比较,指出了属性为连续值时这两种算法的异同及优缺点,其目的是在为解决具体问题时怎样选择这两种算法提供一些有用的线索。

A Comparison between Fuzzy and Crisp Decision Trees

The ID3algorithm is a typical decision tree induction method.A crisp decision tree is based on the precondition that linguistic terms and classes are crisp.Fuzzy decision tree,which is an extension of crisp decision tree,is popularized to deal with the ambiguity and vagueness associated with human thinking.Both of them have been applied to dispose the problem of classification in a wide range.However,there is so far a standard to choose an appropriate one from the two algorithms ,in the real applications.In this paper,we compare the two algorithms based on five aspects,and show their comparative advantages and disadvantages.It aims to provide some useful guidelines for selecting an appropriate algorithm while tree induction is applied to real problems.