Semi-supervised learning on single-view datasets by integration of multiple co-trained classifiers

Proceedings - 2012 11th International Conference on Machine Learning and Applications, ICMLA 2012

Volumn 1, Issue , 2012, Pages 458-463

  Slivka, Jelena ^a   Zhang, Ping ^b   Kovačević, Aleksandar ^a   Konjović, Zora ^a   Obradović, Zoran ^b  

^a UNIVERSITY OF NOVI SAD   (Serbia)

^b TEMPLE UNIVERSITY   (United States)

Author keywords

co training; ensemble methods; multiple annotation; semi supervised learning

Indexed keywords

ALTERNATIVE METHODS; BENCHMARK DATASETS; CLASS LABELS; CO-TRAINING; CO-TRAINING ALGORITHM; CO-TRAINING CLASSIFIERS; ENSEMBLE METHODS; MULTIPLE ANNOTATION; MULTIPLE LABELS; NAIVE BAYES CLASSIFIERS; SEMI-SUPERVISED LEARNING;

LEARNING ALGORITHMS;

SUPERVISED LEARNING;

EID: 84873607124     PISSN: None     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1109/ICMLA.2012.83     Document Type: Conference Paper

References (18)

1. A. Blum and T. Mitchell, "Combining labeled and unlabeled datawith co-training," Proc. 11th Annual Conf. Computational LearningTheory, 92-100, 1998
2. P. Zhang and Z. Obradovic, "Learning from Inconsistent andUnreliable Annotators by a Gaussian Mixture Model and BayesianInformation Criteria," Proc. European Conf. Machine Learning andPrinciples and Practice of Knowledge Discovery in Databases, 553-568, 2011
3. F. Feger and I. Koprinska, "Co-training Using RBF Nets andDifferent Feature Splits," Proc Int'l Joint Conf. Neural Network,1878-1885, 2006
4. J. Du, C. Ling and Z.H. Zhou, "When Does Co-Training Work inReal Data?," IEEE Trans. Knowledge and Data Engineering, 23(5),788-799, 2010
5. K. Nigam and R. Ghani, "Understanding the behavior of co-training,"Proc. Workshop on Text Mining at KDD, 2000
6. K. Nigam and R. Ghani, "Analyzing the Effectiveness andApplicability of Co-training," Proc. Int'l Conf. on Information andKnowledge Management, 86-93, 2000
7. J. Chan, I. Koprinska and J. Poon, "Co-training with a single naturalfeature set applied to email classification," Proc. IEEE/WIC/ACMInt's Conf. Web Intelligence, 586-589, 2004
8. Z.H. Zhou and M. Li, "Tri-training: Exploiting unlabeled data usingthree classifiers," IEEE Trans. Knowledge and Data Engineering, 17,1529-1541, 2005
9. M.F. Hady and F. Schwenker, "Co-training by Committee: A NewSemi-supervised Learning Framework," Proc. IEEE Int'l Conf. DataMining Workshops, 563-572, 2008
10. M. Li and Z.H. Zhou. "Improve computer-aided diagnosis withmachine learning techniques using undiagnosed samples," IEEETrans. Systems, Man and Cybernetics, 37(6), 1088-1098, 2007
11. J. Slivka, A. Kovacevic and Z. Konjovic, "Co-Training BasedAlgorithm for Datasets without the Natural Feature Split," Proc. IEEEInt'l Symp. Intelligent Systems and Informatics, 279-284, 2010
12. I. Muslea, S. Minton and C. Knoblock, "Active + Semi- SupervisedLearning = Robust Multi-View Learning," Proc. Int'l Conf. MachineLearning, 2002
13. J. Huang, J. Sayyad-Shirabad, S. Matwin and J. Su, "Improving Cotrainingwith Agreement-Based Sampling," Proc. Rough Sets andCurrent Trends in Computing, 197-206, 2010
14. G. Salton and C. Buckley, "Term-weighting approaches in automatictext retrieval," Information Processing and Management, 24(5), 513-523, 1988
15. J. Demšar, "Statistical comparisons of classifiers over multiple datasets," Journal of Machine Learning Research, 7:1-30, 2006
16. S. García and F. Herrera, "An extension on statistical comparisons ofclassifiers over multiple data sets for all pairwise comparisons,"Journal of Machine Learning Research, 9:2677-2694, 2008
17. A. Kovaevi, D. Ivanovi, B. Milosavljevi, Z. Konjovi, D. Surla,"Automatic extraction of metadata from scientific publications forCRIS systems," Program: Electronic library and information systems,45(4), 376-396, 2011
18. A. Kovaevi, Z. Konjovi, B. Milosavljevi, G. Nenadic, " Miningmethodologies from NLP publications: A case study in automaticterminology recognition," Computer Speech & Language, 26(2), 105-126, 2011