| 基于数据库分组技术的决策树算法的研究 |
| |
| 引用本文： | 夏平,徐绕山.基于数据库分组技术的决策树算法的研究[J].计算机与数字工程,2007,35(1):34-36. |
| |
| 作者姓名： | 夏平 徐绕山 |
| |
| 作者单位： | 1. 中北大学研究生院,太原,030051
2. 南京信息职业技术学院计算机科学与技术系,南京,210013 |
| |
| 摘 要： | 提出了一种利用数据库分组记数技术构造决策树的算法，利用数据库系统的结构化查询语言来实现主要的样本统计计算任务。并将构造决策树的过程与相关属性的选择方法有机地结合在一起。使用这些方法和策略，该算法能较快速地生成决策树，与ID3算法相比，在不影响决策树分类准确度的前提下，具有较高的执行效率。
|
| 关 键 词： | 数据挖掘 决策树 分组记数 GID3算法 |
| 修稿时间： | 2006年3月10日 |
A Decision Tree Algorithm Based On Database Grouping Technology |
| |
| Xia Ping,Xu Raoshan.A Decision Tree Algorithm Based On Database Grouping Technology[J].Computer and Digital Engineering,2007,35(1):34-36. |
| |
| Authors: | Xia Ping Xu Raoshan |
| |
| Abstract: | This paper studies a classification algorithm that tightly integrates decision tree classification technol- ogy with relational database system technology.In this paper,an algorithm based on grouping and counting is proposed to build the decision tree,which uses the structured query language provided by relational database to implement the major statistics tasks of then training data.Also it binds the feature selection method into the process of building decision tree.With the methods,the algorithm can build a decision tree quickly and efficiently without effect of classification accurency than the tradi- tional algorithm ID3. |
| |
| Keywords: | data mining decision tree grouping and counting algorithm GID_3 |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
