摘要
在研究分类决策中应用得比较成熟和广泛的ID3算法基础上,提出了有效度决策树的模型和Ed算法。Ed算法不但成功解决了ID3算法内在偏置的问题,而且预测精度在某些情况下还会比ID3算法的预测精度高。
Based on investigating the ID3 algorithm which is well-studied and widely used in decision classification methods, proposes a model based on efficiency decision tree, and the ED algorithm. ED algorithm not only settles intrinsic bias of ID3 algorithm, but also improves the precision of prediction better than ID3 algorithm＇s in some situation.
出处
《现代计算机》
2007年第10期31-33,共3页
Modern Computer
关键词
数据挖掘
分类
有效度
Data Mining
Classification
Efficiency
