【摘要】 Boosting是一种有效的分类器组合的方法,文章提出用一个改进的Boosting方法对支持向量机分类器进行集成学习,得到Boosting-MultiSVM分类器;试验结果表明,基于Boosting的支持向量机训练是一个收敛过程,相比标准的支持向量机分类器,Boosting-MultiSVM分类器的泛化性能有不同程度的提高。更多还原

【Abstract】 Boosting is an effective classifier combination method.An improved boosting method is presented for support vector machine classifier combination,and a boosting-MultiSVM classifier is trained.Experimental results show that support vector!machine training based on boosting is convergent.The boosting-MultiSVM classifier has better generalization performance than the standard support vector machine classifier.更多还原