针对训练大样本支持向量机内存开销大、训练速度慢的缺点,提出了一种改进的算法一边界邻近支持向量机。实验表明在分类效果相同情况下,改进算法训练速度明显提高。
Training a support vector machines on a data set of huge size exists one problem with slow training process. We use a modified support vector machines-boundary nearest support vector machines to resolve this problem and it speeds up the training process fastly comparing with conventional support vector machines under the same classification result.