| 使用增量SVM进行文本分类 |
| |
| 引用本文： | 张永,周振龙,侯莉莉,张世宏. 使用增量SVM进行文本分类[J]. 兰州理工大学学报, 2007, 33(1): 100-103 |
| |
| 作者姓名： | 张永 周振龙 侯莉莉 张世宏 |
| |
| 作者单位： | 1. 兰州理工大学,计算机与通信学院,甘肃,兰州,730050
2. 金昌供电公司,甘肃,金昌,737100 |
| |
| 基金项目： | 甘肃省自然科学基金(3ZS042-B25-014) |
| |
| 摘 要： | 针对传统SVM无法适应文本数据库随着时间不断更新的问题,通过对新增文本集的KKT条件的分析,研究了加入新增文本集后支持向量集的变化,提出了使用增量SVM进行文本分类的算法,并通过实验验证了通过该算法得到的分类器和传统分类器有着相似的分类能力和泛化能力.
|
| 关 键 词： | 文本分类 支持向量机 KKT条件 增量 |
| 文章编号： | 1673-5196（2007）01-0100-04 |
| 收稿时间： | 2006-04-06 |
| 修稿时间： | 2006-04-06 |
Text categorization with incremental SVM |
| |
| ZHANG Yong,ZHOU Zhen-long,HOU Li-li,ZHANG Shi-hong. Text categorization with incremental SVM[J]. Journal of Lanzhou University of Technology, 2007, 33(1): 100-103 |
| |
| Authors: | ZHANG Yong ZHOU Zhen-long HOU Li-li ZHANG Shi-hong |
| |
| Abstract: | Aimed at the problem that the traditional support vector machine (SVM) cannot adapt itself to text database updating continuously from time to time, a text categorization algorithm with incremental SVM was presented after analyzing the KKT qualification of newly added text sets and thoroughly investigating the variation of support vector set due to text addition. The experimental results showed that the new algorithm had the same classification and generalization ability as the traditional SVM. |
| |
| Keywords: | text categorization support vector machine KKT qualification increment |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
