摘要
基于SVM寻优问题的KKT条件和样本之间的关系,分析了样本增加后支持向量集的变化情况,支持向量在增量学习中的活动规律,提出了一种新的支持向量机增量学习遗忘机制——计数器淘汰算法。该算法只需设定一个参数,即可对训练数据进行有效的遗忘淘汰。通过对标准数据集的实验结果表明,使用该方法进行增量学习在保证训练精度的同时,能有效地提高训练速度并降低存储空间的占用。
This paper presents a novel approach to incremental Support Vector Machine （SVM） learning algorithm.h analyses the possible change of support vector set after new samples are added to training set.Furthermore,the active properties of SV set in incremental learning are analyzed.After that,a new improved incremental SVM learning algorithm is proposed,which is based on a counter sifting method.The theoretical analysis and experimental results show that this algorithm could not only improve the training speed,but also reduce storage cost.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第6期171-173,共3页
Computer Engineering and Applications
基金
陕西省自然科学基金(the Natural Science Foundation of Shaanxi Province of China under Grant No.2004F36) 。
关键词
支持向量机
增量学习
边界矢量
计数器
support vector machine
incremental learning
margin vector
counter
