| 本文已被：浏览 6780次 下载 7041次 |
码上扫一扫！ |
|
|
| 一种SVM增量学习算法α-ISVM |
|
萧嵘1, 王继成1, 孙正兴1, 张福炎1
|
|
南京大学计算机软件新技术国家重点实验室,江苏,南京,210093,南京大学计算机科学与技术系,江苏,南京,210093
|
|
| 摘要: |
| 基于SVM(support vector machine)理论的分类算法,由于其完善的理论基础和良好的试验结果,目前已逐渐引起国内外研究者的关注.深入分析了SVM理论中SV(support vector,支持向量)集的特点,给出一种简单的SVM增量学习算法.在此基础上,进一步提出了一种基于遗忘因子α的SVM增量学习改进算法α-ISVM.该算法通过在增量学习中逐步积累样本的空间分布知识,使得对样本进行有选择地遗忘成为可能.理论分析和实验结果表明,该算法能在保证分类精度的同时,有效地提高训练速度并降低存储空间的占用. |
| 关键词: 支持向量机 分类 模式识别 增量学习 机器学习 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(69903006,60073030);江苏省"九五"科技重点攻关资助项目(BE96017) |
|
| An Incremental SVM Learning Algorithm α-ISVM |
|
XIAO Rong,WANG Ji-cheng,SUN Zheng-xing,ZHANG Fu-yan
|
| Abstract: |
| The classification algorithm based on SVM (support vector machine) attracts more attention from researchers due to its perfect theoretical properties and good empirical results. In this paper, the properties of SV set are analyzed thoroughly, and a new learning method is introdnced to extend the SVM Classification algorithm to incremental learning area. After that, a new improved incremental SVM learning algorithm is proposed, which is based on a sifting factor. This algorithm accumulates distribution knowledge of the training sample while the incremental training is proceeded, and thus makes it possible to discard samples optimally. The theoretical analysis and experimental results show that this algorithm could not only improve the training speed, but also reduce storage cost. |
| Key words: SVM （support vector machine） classification pattern recognition incremental learning machine learning |
