摘要： 训练支持向量机，可以归结为求解二次规划问题，而求解二次规划时的复杂度随着样本数量的增加而显著增长，这样就大大延长了支持向量机的训练时间。为了提高支持向量机的训练速度，根据支持向量机的基本原理，该文提出了一种从给定训练样本中预抽取支持向量的新方法，即两凸包相对边界向量方法(FFMVM)，此方法大幅度减小了训练支持向量机的训练样本的数量，从而大大提高了支持向量的训练速度，而支持向量机的分类能力不受任何影响。

关键词: 数据挖掘, 支持向量机, 分类, 相对边界向量

Abstract: Training a support vector machine(SVM) can be viewed as a problem of solving a quadratic programming(QP) problem. But the computation complexity of solving QP will increase greatly with the increasing of the amount of training sample, which will increase the training time of SVM largely. In order to improve the speed of training SVM, according to the fundamental principle of SVM, this paper presents a new method called face to face margin vector method of two convex hull which is able to extract support vectors from given training examples for support vector algorithm. This method reduces the training examples largely and improves the speed of training support vector machine greatly, while the ability of support vector machine to classify is unaffected.

中图分类号: 

• TP181