一种基于正区域及其分类纯度的决策树算法

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.