改进的近似支持向量机在葡萄酒质量鉴定中的应用

PSVM-2 Method and Its Application on Identification of Wine Quality

在介绍标准支持向量机(SVM)的基础上,引出近似支持向量机(PSVM)方法的基本原理,并提出一个改进其分类精度的新算法(PSVM-2)。针对葡萄酒质量鉴定这一实际问题,比较了SVM与PSVM及PSVM-2的表现能力,分析了3种算法的复杂性及其分类性能。

Based on introducing standard support vector machine（SVM）,the basic rationale of proximal support vector machine（PSVM）method was led out and a new algorithm（PSVM-2）was proposed to improve its classification precision.For the practical problem of quality assessment of grape wine,the performance abilities of SVM,PSVM and PSVM-2 were compared and their complexities and classification performances were analyzed.