|
|||||
|
|
| 大规模训练集的快速缩减 | |
| 引用本文： | 罗瑜,易文德,何大可,林宇.大规模训练集的快速缩减[J].西南交通大学学报,2007,42(4):468-472,489. |
| 作者姓名： | 罗瑜 易文德 何大可 林宇 |
| 作者单位： | 1. 西南交通大学信息科学与技术学院,四川,成都,610031 2. 重庆文理学院数学与计算机科学系,重庆,402160 3. 西南交通大学经济管理学院,四川,成都,610031 |
| 基金项目： | 上海市特种光纤重点实验科研项目（20050926） |
| 摘 要： | 为了进一步减少支持向量机的训练时间,提出了一种基于类别质心的训练集缩减算法.该算法根据样本的几何分布去除训练集中大部分非支持向量.对样本规模在104数量级的数据集进行了训练实验,结果显示,在基本不损失分类精度的情况下,训练时间比直接用SMO(序贯最小优化)算法减少30%,说明该算法能有效地提高支持向量机的训练速度. |
| 关 键 词： | 支持向量机 类别质心 模式分类 |
| 文章编号： | 0258-2724（2007）04-0468-06 |
| 修稿时间： | 2006-10-23 |
Fast Reduction for Large-Scale Training Data Set |
|
| LUO Yu,YI Wende,HE Dake,LIN Yu.Fast Reduction for Large-Scale Training Data Set[J].Journal of Southwest Jiaotong University,2007,42(4):468-472,489. | |
| Authors: | LUO Yu YI Wende HE Dake LIN Yu |
| Institution: | 1. School of Information Science and Tech. , Southwest Jiaotang University, Chengdu 610031, China; 2. Dept. of Mathematics and Computer Science, Chongqing University of Arts and Sciences, Chongqing 402160, China; 3. School of Economics and Management, Southwest Jiaotong University, Chengdu 610031, China |
| Abstract: | In order to cut down the time of training a large-scale data set by using SVM(support vector machine),a fast algorithm for reducing training sets was proposed based on class centroid.With this algorithm the most of non-support vectors are removed in the light of the geometrical distribution of samples.Experiments were made on several data sets at the level of 104 magnitude.The experimental results show that compared with the SMO(sequential minimal optimization) algorithm,the proposed algorithm decreases training time by 30% under the condition of ensuring the SVM's classification accuracy to greatly improve SVM's training speed. |
| Keywords: | support vector machine class centroid pattern classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |