| 引用本文: | 廖东平,魏玺章,黎湘,等.一种支持向量机增量学习淘汰算法.[J].国防科技大学学报,2007,29(3):65-70.[点击复制] |
| LIAO Dongping,WEI Xizhang,LI Xiang,et al.A Removing Algorithm for Incremental Support Vector Machine Learning[J].Journal of National University of Defense Technology,2007,29(3):65-70[点击复制] |
|
| |
|
|
| 本文已被：浏览 6896次 下载 6671次 |
| 一种支持向量机增量学习淘汰算法 |
| 廖东平, 魏玺章, 黎湘, 庄钊文 |
|
(国防科技大学 电子科学与工程学院，湖南 长沙 410073)
|
| 摘要: |
| 针对大规模数据集的分类问题，支持向量机的训练成为一个难题。增量学习是解决这一难题的思路之一。分析了新增样本加入训练集后支持向量集的变化情况，提出了一种基于密度法的支持向量机增量学习淘汰算法，淘汰了对最终分类无用的样本，在保证测试精度的同时减少了训练时间。实验仿真证明这种算法是有效的。 |
| 关键词: 支持向量机 增量学习 支持向量 |
| DOI： |
| 投稿日期：2006-12-23 |
| 基金项目:国家部委基金资助项目(41303040203) |
|
| A Removing Algorithm for Incremental Support Vector Machine Learning |
| LIAO Dongping, WEI Xizhang, LI Xiang, ZHUANG Zhaowen |
|
(College of Electronic Science and Engineering, National Univ. of Defense Technology, Changsha 410073, China)
|
| Abstract: |
| The training of support vector machine is a difficult issue in classifying large-scale data set. Incremental learning is one of the solutions to the difficulty. After new samples were added to training set, the possible changes of support vector set, were analyzed and a removing algorithm based on density for incremental support vector machine learning was presented. It discarded useless samples, kept the testing accuracy and reduced the training time. Experiments show the validity of this algorithm. |
| Keywords: support vector machine (SVM) incremental learning support vector (SV) |
|
|
|
|
|
