半监督支持向量机及其在县域可持续发展综合实力评价中的应用  

Semi-Supervised Support Vector Machines and Its Application to County Sustainable Development Strength Evaluation

作　　者：杨丽明[1] 党兰芬 

机构地区：[1]中国农业大学理学院,北京100094 [2]苏州大学物理科学与技术学院,江苏苏州215006

出　　处：《数学的实践与认识》2005年第6期17-21,共5页Mathematics in Practice and Theory

摘　　要：给出了基于全部风险(ORM)最小化基础上的半监督支持向量机分类算法,该算法通过加入工作集进行训练,提高了标准SVM对训练集提供信息不充分的数据集的分类泛化能力,而且能有效地处理大量的无标示数据.并将凹半监督支持向量机算法应用于县域可持续发展综合实力评价中.通过邯郸15个县作实证分析,论证了该算法的可行性和有效性.The paper have proposed semi-supervised support vector machines algorithms based on overall risk minmization, which can improve SVM generalization ability when insufficient training information is availiable by incorporating working set into training set and, can efficiently handle large datasets that are mostly unlabeled. The numerical results show that algorithms are efficient and viable by 15 counties of Handan sustainable development strength evaluation.

关 键 词：半监督支持向量机 凹规划 数据分类 可持续发展 数据挖掘 

分 类 号：F224[经济管理—国民经济]