| 基于直推式支持向量机的图像分类算法 |
| |
| 引用本文： | 沈新宇,许宏丽,官腾飞.基于直推式支持向量机的图像分类算法[J].计算机应用,2007,27(6):1463-1464. |
| |
| 作者姓名： | 沈新宇 许宏丽 官腾飞 |
| |
| 作者单位： | 北京交通大学,计算机与信息技术学院,北京,100044 |
| |
| 基金项目： | 国家自然科学基金
,
北京交通大学校科研和教改项目 |
| |
| 摘 要： | 直推式支持向量机(TSVM)是在利用有标签样本的同时，考虑无标签样本对分类器的影响，并且结合支持向量机算法，实现一种高效的分类算法。它在包含少量有标签样本的训练集和大量无标签样本的测试集上，具有良好的效果。但是它有算法时间复杂度比较高，需要预先设置正负例比例等不足。通过对原有算法的改进，新算法在时间复杂度上明显下降，同时算法效果没有明显的影响。
|
| 关 键 词： | 支持向量机 直推式学习 图像分类 |
| 文章编号： | 1001-9081（2007）06-1463-02 |
| 收稿时间： | 2006-12-15 |
| 修稿时间： | 2006-12-152007-02-26 |
Image classification based on transductive support vector machines |
| |
| SHEN Xin-yu,XU Hong-li,GUAN Teng-fei.Image classification based on transductive support vector machines[J].journal of Computer Applications,2007,27(6):1463-1464. |
| |
| Authors: | SHEN Xin-yu XU Hong-li GUAN Teng-fei |
| |
| Affiliation: | Computer Information and Technology Institute, Beijing Jiaotong University, Beijing 100044, China |
| |
| Abstract: | Transductive Support Vector Machines (TSVM) take advantage of the test sets as well as the train sets and inherit most properties of inductive SVMs. They are more efficient than inductive SVMs, especially for very small training sets and large test sets. But they still have disadvantages, such as high time complexity and the requirement of "num ". The improved algorithm substantially reduces the time complexity with little influence on the performance. |
| |
| Keywords: | Support Vector Machine (SVM) transductive learning image classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机应用》浏览原始摘要信息 |
|
点击此处可从《计算机应用》下载全文 |
