Dynamic Conditional Random Fields: Factorized Probabilistic Models for Labeling and Segmenting Sequence Data
Abstract
In sequence modeling, we often wish to represent complex interaction between labels, such as when performing multiple, cascaded labeling tasks on the same sequence, or when long-range dependencies exist. We present {\em dynamic conditional random fields (DCRFs)}, a generalization of linear-chain conditional random fields (CRFs) in which each time slice contains a set of state variables and edges---a distributed state representation as in dynamic Bayesian networks (DBNs)---and parameters are tied across slices. Since exact inference can be intractable in such models, we perform approximate inference using several schedules for belief propagation, including tree-based reparameterization (TRP). On a natural-language chunking task, we show that a DCRF performs better than a series of linear-chain CRFs, achieving comparable performance using only half the training data.
Cite
Text
Sutton et al. "Dynamic Conditional Random Fields: Factorized Probabilistic Models for Labeling and Segmenting Sequence Data." International Conference on Machine Learning, 2004. doi:10.1145/1015330.1015422Markdown
[Sutton et al. "Dynamic Conditional Random Fields: Factorized Probabilistic Models for Labeling and Segmenting Sequence Data." International Conference on Machine Learning, 2004.](https://mlanthology.org/icml/2004/sutton2004icml-dynamic/) doi:10.1145/1015330.1015422BibTeX
@inproceedings{sutton2004icml-dynamic,
title = {{Dynamic Conditional Random Fields: Factorized Probabilistic Models for Labeling and Segmenting Sequence Data}},
author = {Sutton, Charles and Rohanimanesh, Khashayar and McCallum, Andrew},
booktitle = {International Conference on Machine Learning},
year = {2004},
doi = {10.1145/1015330.1015422},
url = {https://mlanthology.org/icml/2004/sutton2004icml-dynamic/}
}