针对大规模训练集的支持向量机的学习策略

A Learning Strategy of SVM Used to Large Training Set

当训练集的规模很大特别是支持向量很多时 ,支持向量机的学习过程需要占用大量的内存 ,寻优速度非常缓慢 ,这给实际应用带来了很大的麻烦 .该文提出了一种针对大规模样本集的学习策略 :首先用一个小规模的样本集训练得到一个初始的分类器 ,然后用这个分类器对大规模训练集进行修剪 ,修剪后得到一个规模很小的约减集 ,再用这个约减集进行训练得到最终的分类器 .实验表明 ,采用这种学习策略不仅大幅降低了学习的代价 ,而且这样获得的分类器的分类精度完全可以与直接通过大规模样本集训练得到的分类器的分类精度相媲美 ,甚至更优 ,同时分类速度也得到大幅提高 .

This paper proposes a learning strategy of SVM used to large training set. First authors train an initial classifier with a small training set, then prune the large training set with the initial classifier to obtain a small reduction set. Training with the reduction set, final classifier is obtained. Experiments show that the learning strategy not only reduces the cost greatly but also obtains a classifier that has the same accuracy as(even better than) the classifier obtained by training large set directly. In addition, speed of classification is greatly improved.