Abstract: This paper presents a survey of kernel-based support vector machines.The main ideas of support vector machines and kernel feature spaces are described briefly，and then a description of the kernel methods for support vector machines is indroduced，including details of the recent advanced techniques and its key applications.

Key words: kernel, support vector machines, generalization performance

摘要： 概述了基于核函数方法的支持向量机。首先简要叙述支持向量机的基本思想和核特征空间，然后重点介绍核函数支持向量机的前沿理论与领先技术，同时描述了核函数支持向量机在关键领域的应用。

关键词: 核函数, 支持向量机, 推广性能