
Unpublished Paper
A Note on Semi-Supervised Learning using Markov Random Fields
(2004)
Abstract
This paper describes conditional-probability training of Markov random fields using combinations of labeled and unlabeled data. We capture the similarities between instances learning the appropriate distance metric from the data. The likelihood model and several training procedures are presented.
Disciplines
Publication Date
2004
Comments
This is the pre-published version harvested from CIIR.
Citation Information
Wei Li and Andrew McCallum. "A Note on Semi-Supervised Learning using Markov Random Fields" (2004) Available at: http://works.bepress.com/andrew_mccallum/95/