Co-EM Support Vector learning

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Authors

Multi-view algorithms, such as co-training and co-EM, utilize unlabeled data when the available attributes can be split into independent and compatible subsets. Co-EM outperforms co-training for many problems, but it requires the underlying learner to estimate class probabilities, and to learn from probabilistically labeled data. Therefore, co-EM has so far only been studied with naive Bayesian learners. We cast linear classifiers into a probabilistic framework and develop a co-EM version of the Support Vector Machine. We conduct experiments on text classification problems and compare the family of semi-supervised support vector algorithms under different conditions, including violations of the assumptions underlying multiview learning. For some problems, such as course web page classification, we observe the most accurate results reported so far.

Original languageEnglish
Title of host publicationProceeding ICML '04 Proceedings of the twenty-first international conference on Machine learning
Number of pages8
Place of PublicationNew York
PublisherAssociation for Computing Machinery, Inc
Publication date2004
Pages121-128
ISBN (Print)1-58113-838-5 , 978-1-58113-838-2
DOIs
Publication statusPublished - 2004
Event21st International Conference on Machine Learning - 2004 - Banff, Canada
Duration: 31.12.2004 → …
Conference number: 21
https://icml.cc/imls/icml.html

DOI