Training Conditional Random Fields using Virtual Evidence Boosting
Submitted by mkrainin on Tue, 2012-02-28 12:21
| Title | Training Conditional Random Fields using Virtual Evidence Boosting |
| Publication Type | Conference Paper |
| Year of Publication | 2007 |
| Authors | Liao L, Choudhury T, Fox D, Kautz H |
| Conference Name | IJCAI |
| Abstract | <p>While conditional random ï¬elds (CRFs) have been applied successfully in a variety of domains, their training remains a challenging task. In this paper, we introduce a novel training method for CRFs, called virtual evidence boosting, which simulta- neously performs feature selection and parameter estimation. To achieve this, we extend standard boosting to handle virtual evidence, where an ob- servation can be speciï¬ed as a distribution rather than a single number. This extension allows us to develop a uniï¬ed framework for learning both local and compatibility features in CRFs. In experiments on synthetic data as well as real activity classiï¬- cation problems, our new training algorithm out- performs other training approaches including max- imum likelihood, maximum pseudo-likelihood, and the most recent boosted random ï¬elds.</p> |
| Downloads | |
| Citation Key | Lia07Tra |
Last changed Tue, 2012-03-20 17:58

cs.