一种新的支持向量机增量学习算法

支持向量机学习算法的本质是从训练集中寻找支持向量,因此能否通过训练算法能快速找出支持向量是衡量支持向量机算法优劣的重要标准.本文提出了一种新的快速训练支持向量机的增量学习算法,首先,给出边界向量的定义,然后,对一个给定的新加人的样本,新的学习方法验证其是否为边界向量,如果是,将其加入到训练集中重新训练支持向量机,如果不是,就舍弃,这样能达到减少训练样本、降低训练复杂性目的,最后,给出了一个增量学习算法.实验表明测试误差和支持向量数量与SMO算法大致相当,而训练速度明显加快.

A New Incremental Learning for Support Vector Machine

A new incremental learning method for support vector machine(SVM)is proposed,which trains SVM quickly and incrementally.In this paper,a definition of border vectors is given first.Border vectors are supported by vector candidates.Then for a given new-added sample,the proposed training method validates whether they are border vectors or not.If true,they are added to training sample set to retrain support vector machine,otherwise omitted.Hence,the training samples can be reduced and training complexity can be...