模式识别与人工智能
2025年4月2日 星期三   首 页     期刊简介     编委会     投稿指南     伦理声明     联系我们                                                                English
模式识别与人工智能  2016, Vol. 29 Issue (1): 39-46    DOI: 10.16451/j.cnki.issn1003-6059.201601005
论文与报告 最新目录| 下期目录| 过刊浏览| 高级检索 |
结合Tri-training半监督学习和凸壳向量的SVM主动学习算法*
徐海龙,龙光正,别晓峰,吴天爱,郭蓬松
空军工程大学 防空反导学院 西安 710051
Active Learning Algorithm of SVM Combining Tri-training Semi-supervised Learning and Convex-Hull Vector
XU Hailong, LONG Guangzheng, BIE Xiaofeng, WU Tian′ai, GUO Pengsong
Air and Missile Defense College, Air Force Engineering University, Xi'an 710051

全文: PDF (556 KB)   HTML (1 KB) 
输出: BibTeX | EndNote (RIS)      
摘要 为解决监督学习过程中难以获得大量带有类标记样本且样本数据标记代价较高的问题,结合主动学习和半监督学习方法,提出基于Tri-training半监督学习和凸壳向量的SVM主动学习算法.通过计算样本集的壳向量,选择最有可能成为支持向量的壳向量进行标记.为解决以往主动学习算法在选择最富有信息量的样本标记后,不再进一步利用未标记样本的问题,将Tri-training半监督学习方法引入SVM主动学习过程,选择类标记置信度高的未标记样本加入训练样本集,利用未标记样本集中有利于学习器的信息.在UCI数据集上的实验表明,文中算法在标记样本较少时获得分类准确率较高和泛化性能较好的SVM分类器,降低SVM训练学习的样本标记代价.
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
徐海龙
龙光正
别晓峰
吴天爱
郭蓬松
关键词 主动学习 半监督学习 支持向量机(SVM) 凸壳向量 Tri-training算法    
Abstract:The large-scale labeled samples can not be acquired easily and the cost of sample labeling is high. Aiming at these problems, an active learning algorithm of support vector machine (SVM) based on tri-training semi-supervised learning and convex-hull vector is proposed in this paper. Semi-supervised learning and active learning are efficiently combined. Firstly, by calculating the convex-hull vector of the sample set, samples of convex-hull vector which are most likely to be support vectors are selected to be labeled. For the existing active learning, the unlabeled samples are no longer used after the most informative samples are selected to be labeled. Secondly, to salve this problem, semi-supervised learning method-based tri-training is introduced into SVM active learning. Thus, the unlabeled samples with higher confidence level of classifying samples are selected and classified as the training sample set, and the useful information for learning machines in the unlabeled samples is exploited. The experimental results on UCI dataset show that the proposed algorithm achieves higher classification accuracy with less labeled samples and it improves generalization performance and reduces the labeling cost of SVM training.
Key wordsActive Learning    Semi-supervised Learning    Support Vector Machine (SVM)    Convex-Hull Vector    Tri-training Algorithm   
收稿日期: 2015-01-29     
ZTFLH: TP 181  
基金资助:国家自然科学基金项目(No.61273275)资助
作者简介: 徐海龙(通讯作者),男,1981年生,博士,讲师,主要研究方向为模式识别、支持向量机、目标分配等.E-mail:xhl_81329@163.com.龙光正,男,1975年生,博士,副教授,主要研究方向为目标识别、分配、作战仿真等.E-mail:lunggz@163.com.别晓峰,男,1979年生,博士,副教授,主要研究方向为目标分配、作战评估与仿真等.E-mail:58719591@qq.com.吴天爱,女,1977年生,博士研究生,讲师,主要研究方向为目标跟踪与识别.E-mail:wuyh7277@163.com.郭蓬松,男,1979年生,硕士,副教授,主要研究方向为目标分配、作战评估与仿真.E-mail:xupengss@163.com.
引用本文:   
徐海龙,龙光正,别晓峰,吴天爱,郭蓬松. 结合Tri-training半监督学习和凸壳向量的SVM主动学习算法*[J]. 模式识别与人工智能, 2016, 29(1): 39-46. XU Hailong, LONG Guangzheng, BIE Xiaofeng, WU Tian'ai, GUO Pengsong. Active Learning Algorithm of SVM Combining Tri-training Semi-supervised Learning and Convex-Hull Vector. , 2016, 29(1): 39-46.
链接本文:  
http://manu46.magtech.com.cn/Jweb_prai/CN/10.16451/j.cnki.issn1003-6059.201601005      或     http://manu46.magtech.com.cn/Jweb_prai/CN/Y2016/V29/I1/39
版权所有 © 《模式识别与人工智能》编辑部
地址:安微省合肥市蜀山湖路350号 电话:0551-65591176 传真:0551-65591176 Email:bjb@iim.ac.cn
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn