Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Chinese Journal of Computers》 2007-08
Add to Favorite Get Latest Update

ADE-Tri-training:Tri-training with Adaptive Data Editing

DENG Chao GUO Mao-Zu(School of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001)  
Tri-training, a Co-training style semi-supervised learning algorithm, can effectively exploit unlabeled examples to improve generalization ability. However, Tri-training may suffer more from the common problem in semi-supervised learning, i.e. the performance is usually not stable due to the unlabeled examples may often be wrongly labeled and accumulated during the iterative learning process. In this paper a new Tri-training style algorithm named ADE-Tri-training (Tri-training with Adaptive Data Editing) is proposed. ADE-Tri-training not only employs a specific Data Editing technique to identify and discard possible mislabeled examples along with iterations of three classifiers mutually labeling, but also takes an adaptive strategy to trigger or inhibit the editing operation according to different situation. The adaptive strategy is combinations of five precondition theorems all that will ensure reducing classification error as well as increasing the scale of new training set iteratively under the PAC theory. This paper also provides the proof of all these precondition theorems. Experiments on UCI datasets show that ADE-Tri-training could more effectively and stably utilize the unlabeled examples to improve classification generalization than Tri-training and DE-Tri-training (Tri-training with Data Editing but without adaptive strategy).
【Key Words】: semi-supervised learning data editing adaptive strategy PAC learning Tri-training
【Fund】: 国家自然科学基金(60671011);; 黑龙江省杰出青年科学基金(JC200611);; 黑龙江省留学回国人员科技项目;; 哈尔滨工业大学校基金(HIT.2003.53)资助~~
【CateGory Index】: TP181
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
【References】
Chinese Journal Full-text Database 2 Hits
1 WANG Jiao,LUO Si-wei,ZENG Xian-hua(School of Computer and Information Technology,Beijing Jiaotong University,Beijing 100044,China);A Random Subspace Method for Co-Training[J];Acta Electronica Sinica;2008-S1
2 LI Kun-lun,ZHANG Wei,DAI Yun-na College of Electronic and Information Engineering,Hebei University,Baoding,Hebei 071002,China;Semi-supervised SVM based on Tri-training[J];Computer Engineering and Applications;2009-22
【Co-references】
Chinese Journal Full-text Database 10 Hits
1 CAI Yingkun XIE Kunqing MA Xiujun (Center of Information Science, Peking University, Beijing, 100871);An Improved DBSCAN Algorithm which is Insensitive to Input Parameters[J];Acta Scicentiarum Naturalum Universitis Pekinesis;2004-03
2 FENG Jufu SHI Jianxin (Center for Information, National Laboratory on Machine Perception, School of Electronics Engineering and Computer Science, Peking University, Beijing,100871);Gene Selection Based on Fast Fisher Optimization Model[J];Acta Scicentiarum Naturalum Universitis Pekinesis;2005-01
3 WU Zhi-feng,TIAN Xue-dong (College of Mathematics and Computer,Hebei University,Baoding 071002,China);Application of Name of People and Institution in Text Categorization[J];Journal of Hebei University(Natural Science Edition);2004-06
4 Zhu Hong Wu Lin Zhu Hong Assoc. Prof.; College of Computer Sci. & Tech., Huazhong Univ. of Sci. & Tech., Wuhan 430074, China.;Compression of inverted and implementation in full-text information retrieval system RDBMS[J];Journal of Huazhong University of Science and Technology;2005-04
5 ZHOU Shui Geng, ZHOU Ao Ying, and CAO Jing (Department of Computer Science, Fudan University, Shanghai 200433) (Shanghai (International) Database Research Center, Shanghai 200433);A DATA-PARTITIONING-BASED DBSCAN ALGORITHM[J];JOURNAL OF COMPUTER RESEARCH AND DEVELOPMENT;2000-10
6 ZENG Hai Quan, LIU Yong Dan, Song Yang, HU Yun Fa (Department of Computer and Information Technology, Fudan University, Shanghai 200433);Mining Relationship Patterns in Multiple Time Series Based on IRST[J];Journal of Computer Research and Development;2003-07
7 LI Rong Lu and HU Yun Fa (Department of Computing and Information Technology, Fudan University, Shanghai 200433);A Density-Based Method for Reducing the Amount of Training Data in kNN Text Classification[J];Journal of Computer Research and Development;2004-04
8 ZHANG Meng, WANG Da Ling, and YU Ge (Institute of Information Science and Engineering, North Eastern University, Shenyang 110004);A Text Clustering Method Based on Auto-Selected Threshold[J];Journal of Computer Research and Development;2004-10
9 Wang Jianhui , Wang Hongwei , Shen Zhan, and Hu Yunfa(Department of Computing and Information Technology, Fudan University, Shanghai 200433) (School of Economics and Management, Tongji University, Shanghai 20433);A Simple and Efficient Algorithm to Classify a Large Scale of Texts[J];Journal of Computer Research and Development;2005-01
10 Li Ronglu, Wang Jianhui, Chen Xiaoyun, Tao Xiaopeng, and Hu Yunfa (Department of Computing and Information Technology, Fudan University, Shanghai 200433);Using Maximum Entropy Model for Chinese Text Categorization[J];Journal of Computer Research and Development;2005-01
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved