【摘要】 论述基于核函数的方法,包括支持向量机的分类、聚类与回归算法,基于核的Fisher判别分析、感知机和主成分分析,介绍学习算法、简化算法和多类分类等研究课题,及基于核函数方法的应用.更多还原

【Abstract】 This paper describes the kernelbased methods including the classification, clustering and regression of support vector machines, Fisher discriminant analysis, perceptron and principal component analysis based on kernel. The research topics including learning algorithms, simplification algorithms and multiclassese classification, and the applications of kernel_based methods are discussed.更多还原