文献类型：期刊文献

中文题名：核函数方法及其模型选择

英文题名：Study on the KernelBased Methods and Its Model Selection

作者：王华忠[1];俞金寿[1]

机构：[1]华东理工大学自动化研究所,上海200237

年份：2006

卷号：5

期号：4

起止页码：500

中文期刊名：江南大学学报（自然科学版）

外文期刊名：Joural of Jiangnan University (Natural Science Edition)　

收录：CSTPCD

语种：中文

中文关键词：核函数方法;支持向量机;模型选择;核主元分析

外文关键词：kernel-based methods; support vector machines（SVMs）; model selection; kernel principal component analysis

摘要：核函数方法已成为近年来机器学习领域继人工神经网络方法之后又一个十分流行和有效的方法.阐述了核函数方法的基本原理、特点及实施步骤,介绍了几种主要的核函数方法,最后重点分析和讨论了核函数方法中参数选择和核函数构造等核函数方法研究中的热点问题,并对其未来研究作了展望.
Kernebbased methods have become another very popular and important techniques after artificial neural networks in the field of machine learning. The principles characteristics and implementation of kernel-based methods are introduced and several important kernel-based methods are briefly described. Hot issues such as model selection, construction of kernels are discussed in detail and future research is proposed.