| 预抽取相对较近边界向量的选块算法 |
| |
| 引用本文： | 孔波,刘小茂,曹淑娟,苏展.预抽取相对较近边界向量的选块算法[J].计算机工程与应用,2006,42(28):170-173. |
| |
| 作者姓名： | 孔波 刘小茂 曹淑娟 苏展 |
| |
| 作者单位： | 华中科技大学主校区数学系,武汉,430074 |
| |
| 基金项目： | 国家自然科学基金;航空基础科学基金 |
| |
| 摘 要： | 利用支持向量机中支持向量的稀疏性和支持向量分布于分划超平面周围的性质,该文提出了一种预抽取相对较近边界向量的选块算法的新算法,该算法减少了普通选块算法的迭代次数和提高了仅依靠相对较近边界向量的准确率,从而大大加快了支持向量机的训练速度,且支持向量机的分类能力不受任何影响。
|
| 关 键 词： | 支持向量机 相对较近边界向量 选块算法 |
| 文章编号： | 1002-8331（2006）28-0170-04 |
| 收稿时间： | 2005-12 |
| 修稿时间： | 2005-12 |
Pre-extracting Relatively Closer Margin Vectors for Chunking Algorithm |
| |
| KONG Bo,LIU Xiao-mao,CAO Shu-juan,SU Zhan.Pre-extracting Relatively Closer Margin Vectors for Chunking Algorithm[J].Computer Engineering and Applications,2006,42(28):170-173. |
| |
| Authors: | KONG Bo LIU Xiao-mao CAO Shu-juan SU Zhan |
| |
| Affiliation: | Math Department of Huazhong University of Science and Technology,Wuhan 430074 |
| |
| Abstract: | According to the sparsity of support vectors in Support Vector Machine(SVM) and the fact that support vectors are distributed around the separating hyperplane,this paper presents a new algorithm called Pre-extracting Relatively Closer Margin Vectors for Chunking Algorithm,which reduces the iterative times of Chunking Algorithm and improves the correct rate of classifiers which is just based on the Pre-extracting Relatively Closer Margin Vectors.So the new algorithm improves the speed of SVM greatly,while the ability of SVM to classify is unaffected. |
| |
| Keywords: | Support Vector Machine relatively closer margin vectors chunking algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
