一种改进的SVM算法在决策分析中的应用  

作　　者：孙蕾[1] 温有奎[1] 

机构地区：[1]西安电子科技大学经济管理学院,陕西710071

出　　处：《情报理论与实践》2006年第6期758-760,共3页Information Studies:Theory & Application

基　　金：国家自然科学基金资助项目;项目编号:70373046

摘　　要：决策支持系统是跨学科的综合体系,涉及机器学习理论。支持向量机是近几年发展起来的学习方法,它是利用最优分类面(线)将两类样本在特征空间或输入空间中无错误地分开,而且要使两类的分类空隙最大。然而当两类中的样本数量差别悬殊时,支持向量机的分类能力会下降。为了解决此问题,本文提出了一种改进的支持向量机算法,在所开发的医学决策支持系统上的应用表明,此方法在解决两类样本数量十分不均衡问题时有着很强的分类能力,不失为一种有效的决策分析工具。The decision support system is a cross-disciplinary integrated system, which involves with machine learning. The Support Vector Machine （SVM） is an algorithm of machine learning developed in recent years. SVM is used to correctly classify samples into two parallel planes in input or feature space by optimal planes （lines）. And the margin between the two classes is made to be the largest. However, when the two-class problem samples are very unbalanced, SVM has a poor performance. To solve this problem, an improved SVM is presented. The result on a medical decision support system demonstrates that the modified algorithm has a strong capability of classification for the unbalanced samples of the two-class problems. And it is an effective tool of decision analysis.

关 键 词：决策支持系统 机器学习 支持向量机 分类算法 决策分析 SVM算法 

分 类 号：C934[经济管理—管理学]