首页 | 本学科首页   官方微博 | 高级检索  
     

一种改进的DDAGSVM决策算法
引用本文:史朝辉,王坚,华继学,郭新鹏. 一种改进的DDAGSVM决策算法[J]. 空军工程大学学报(自然科学版), 2015, 16(2): 53-56
作者姓名:史朝辉  王坚  华继学  郭新鹏
作者单位:空军工程大学防空反导学院,西安,710051
基金项目:国家自然科学基金资助项目
摘    要:决策导向无环图支持向量机(DDAGSVM)是一种典型的SVM多类分类算法,然而传统SVM决策分类器存在误差积累,其推广能力有待进一步提高。为改进DDAGSVM,有效的做法是定义一种类间可分离性测度,将容易分的类先分割出来,然后再分不容易分的类,使错分尽可能地远离图的根部。引入了一种基于广义KKT条件的类间可分离性测度,提出一种改进的DDAGSVM分类决策算法。三螺旋线实验和HRRP分类实验证明该方法对控制分类错误有明显的效果。

关 键 词:支持向量机  决策导向无环图  广义KKT条件  可分离性测度

An Improved Algorithm for DDAGSVM
SHI Zhao-hui,WANG Jian,HUA Ji-xue,GUO Xin-peng. An Improved Algorithm for DDAGSVM[J]. Journal of Air Force Engineering University(Natural Science Edition), 2015, 16(2): 53-56
Authors:SHI Zhao-hui  WANG Jian  HUA Ji-xue  GUO Xin-peng
Affiliation:SHI Zhao-hui;WANG Jian;HUA Ji-xue;GUO Xin-peng;Air and Missile Defense College,Air Force Engineering University;
Abstract:A decision directed acyclic graph support vector machine is a typical multi-class classification with support vector machines. But error accumulation exists in the traditional decision classification, and its generalization ability depends on the tree structure. In this paper, to improve the generalization ability of DDAGSVM, a novel separable measure is defined based on the generalized KKT, and an improved decision directed acyclic graph support vector machine is given. The three-spiral and HRRP experimental results show that this kind of algorithm has an obvious effectiveness in controlling classification errors.
Keywords:support vector machine  decision directed acyclic graph  generalized KKT  separable measure
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《空军工程大学学报(自然科学版)》浏览原始摘要信息
点击此处可从《空军工程大学学报(自然科学版)》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号