| 基于自适应距离度量的分类器设计方法 |
| |
| 引用本文： | 郭亚琴,王正群,乐晓蓉,王向东. 基于自适应距离度量的分类器设计方法[J]. 计算机工程与设计, 2007, 28(10): 2270-2272 |
| |
| 作者姓名： | 郭亚琴 王正群 乐晓蓉 王向东 |
| |
| 作者单位： | 扬州大学,信息工程学院,江苏,扬州,225009;扬州大学,信息工程学院,江苏,扬州,225009;扬州大学,信息工程学院,江苏,扬州,225009;扬州大学,信息工程学院,江苏,扬州,225009 |
| |
| 基金项目： | 江苏省高校自然科学基金 , 扬州大学校科研和教改项目 |
| |
| 摘 要： | 通过对欧氏距离度量的分析,提出了自适应距离度量.首先利用训练样本建立自适应距离度量模型,该模型保证了训练样本到相同模式类的距离最近,到不同模式类的距离最远,根据该模型建立目标函数,求解目标函数,得到最优权重.基于最小距离分类器和K近邻分类器,采用UCI标准数据库中部分数据,对提出的自适应距离度量和欧氏距离度量进行了实验比较,实验结果表明自适应距离度量更有效.
|
| 关 键 词： | 分类 最小距离分类器 K 近邻分类器 自适应距离度量 最优权重 |
| 文章编号： | 1000-7024（2007）10-2270-03 |
| 修稿时间： | 2006-05-02 |
Classifier design based on adaptive distance metric |
| |
| GUO Ya-qin,WANG Zheng-qun,LE Xiao-rong,WANG Xiang-dong. Classifier design based on adaptive distance metric[J]. Computer Engineering and Design, 2007, 28(10): 2270-2272 |
| |
| Authors: | GUO Ya-qin WANG Zheng-qun LE Xiao-rong WANG Xiang-dong |
| |
| Affiliation: | School of Information Engineering, Yangzhou University, Yangzhou 225009, China |
| |
| Abstract: | After the Euclidian distance metric is analyzed,a new training method based on adaptive distance metric is proposed,which adopts a model about adaptive distance metric,the model assures that the distance between training sample and the same pattern classi-fication is nearest,and the distance between training sample and other pattern classification is far,then a optimal weight is obtained through solving objective function.By adding weight define for distance in the classification phase,the classifier improved its classifi-cation accuracy.Experiment is tested on UCI standard database,the results show that the proposed minimum distance classifier and K-nearest neighbor classifier based on adaptive distance metric is effective,and it is superior to Euclidian distance metric in classification performance. |
| |
| Keywords: | classification minimum distance classifier k-nearest classifier adaptive distance metric optimal weight |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
