基于取样的潜在支持向量机序列最小优化算法

A Sequential Minimal Optimization Algorithm for the Potential Support Vector Machine Based on Sampling

为了提高潜在支持向量机求解大规模问题的训练速度,提出了基于样本取样的潜在支持向量机序列最小优化算法,去掉了大部分非支持向量,把支持向量逐渐压缩到取样样本集中.此算法特别适合大样本数据且支持向量个数相对较少的情况.实验表明,改进的序列最小优化算法加速了潜在支持向量机分类器训练时间.

To accelerate the training speed of the Potential Support Vector Machine（PSVM）for large-scale datasets,a new method is proposed,which introduces the sequential minimal optimization（SMO）algorithm based on sampling for PSVM.The new method removes most non-support vectors,and compresses the support vectors to the sampling set.This method is more suitable for large-scale datasets with relatively small number of support vectors.The experimental results show that the improved SMO algorithm decreases the training time.