一种新型回归支持向量机的学习算法

支持向量机(SVM)是一种基于结构风险最小化原理的学习技术,也是一种具有很好泛化性能的回归方法,本文对标准支持向量机稍作改动,提出了一种新型回归支持向量机,并推导出它的对偶表达方式,随后利用一个优化定理设计了一个多变量更新学习算法,该算法能单调收敛于极值点,并具有简单的迭代方式,仿真实例说明所提出的回归支持向量机及其训练算法具有较好的学习性能.

Learning Algorithm for a New Regression SVM

Support vector machine(SVM) is a learning technique based on the structural risk minimization principle,and it is also a class of regression method with good generalization ability.By modifying the standard SVM formulation,this paper proposes a new kind of regression SVM and deduces its dual formulations.Then a multiplicative updates learning algorithm is designed on the basis of an optimal theorem.The updates algorithm converges monotonically on the extremum solution of the optimal problem and has a simple iterative form.Experimental results of simulation indicate the feasibility of the new regression SVM and its training algorithm.