详细信息
文献类型:期刊文献
中文题名:非线性支持向量分类机解的二阶充分条件
英文题名:Second Order Sufficient Conditions Property for Support Vector Classification
作者:蔡春[1];吕书强[1];徐坚[2]
第一作者:蔡春
机构:[1]北京联合大学应用文理学院;[2]北京联合大学应用科技学院
第一机构:北京联合大学应用文理学院
年份:2015
卷号:0
期号:3
起止页码:264-267
中文期刊名:数学的实践与认识
外文期刊名:Mathematics in Practice and Theory
收录:CSTPCD;;北大核心:【北大核心2014】;
基金:北京市属高等学校高层次人才引进与培养计划项目;CIT&TCD201404080;北京联合大学新起点计划项目资助~~
语种:中文
中文关键词:支持向量分类机;二阶充分条件;灵敏度分析;数据挖掘
外文关键词:support vector classification; second order sufficient condition; sensitivity analysis; data mining
摘要:优化问题解的二阶充分条件是研究其灵敏度分析的基础,支持向量分类机是新的数据挖掘优化问题.给出了支持向量分类机的解满足二阶充分条件成立定理;定理的假设条件是很弱的,用支持向量分类机求解实际问题,通常总假定这一条件成立;特别地,对线性可分支持向量机问题,其解满足二阶充分条件成为当然成立的事实.
Second order sufficient condition is the foundation for optimal problem sensitivity analysis. Support vector classification is new data mining optimal model. The paper presents one hypothesis which guarantees that second order sufficient condition property of support vector classification holds. The hypothesis is much weak. One reason is that the hypothesis holds certainly for linearly separable support vector classification. In addition, the linear support vector classification solution is solved under such a hypothesis.
参考文献:
正在载入数据...