The improved C4.5 classifier and its application in recognition of moving targets

被引：0

作者：

Yao, Yafu ^{[1
]}

Xing, Liutao ^{[1
]}

机构：

[1] Cent S Univ, Sch Mech & Elect Engn, Changsha 410083, Peoples R China

来源：

2011 INTERNATIONAL CONFERENCE ON COMPUTERS, COMMUNICATIONS, CONTROL AND AUTOMATION (CCCA 2011), VOL I | 2010年

关键词：

decision tree; C4.5; classifier; K-mean classifier; target identification; information gain ratio;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

According to the complexity of calculating the information gain ratio in traditional C4.5 algorithm, an improved method of calculating information gain ratio is proposed in this paper. An improved C4.5 classifier which adopts the improved method of calculating information gain ratio is established and trained in this paper, and then it is applied in the recognition of people and vehicle targets in video sequences. The results of experiment show that the improved C4.5 classifier greatly improves the efficiency of generating a decision tree, at the same time, the classification accuracy to some extent is increased, and the improved information gain ratio algorithm has good prospects and potential economic value.

引用

页码：572 / 577

页数：6