TY - GEN
T1 - CRF-OPT
T2 - 23rd AAAI Conference on Artificial Intelligence and the 20th Innovative Applications of Artificial Intelligence Conference, AAAI-08/IAAI-08
AU - Chen, Minmin
AU - Chen, Yixin
AU - Brent, Michael R.
PY - 2008
Y1 - 2008
N2 - Conditional random field (CRF) is a popular graphical model for sequence labeling. The flexibility of CRF poses significant computational challenges for training. Using existing optimization packages often leads to long training time and unsatisfactory results. In this paper, we develop CRF-OPT, a general CRF training package, to improve the efficiency and quality for training CRFs. We propose two improved versions of the forward-backward algorithm that exploit redundancy and reduce the time by several orders of magnitudes. Further, we propose an exponential transformation that enforces sufficient step sizes for quasi-Newton methods. The technique improves the convergence quality, leading to better training results. We evaluate CRF-OPT on a gene prediction task on pathogenic DNA sequences, and show that it is faster and achieves better prediction accuracy than both the HMM models and the original CRF model without exponential transformation.
AB - Conditional random field (CRF) is a popular graphical model for sequence labeling. The flexibility of CRF poses significant computational challenges for training. Using existing optimization packages often leads to long training time and unsatisfactory results. In this paper, we develop CRF-OPT, a general CRF training package, to improve the efficiency and quality for training CRFs. We propose two improved versions of the forward-backward algorithm that exploit redundancy and reduce the time by several orders of magnitudes. Further, we propose an exponential transformation that enforces sufficient step sizes for quasi-Newton methods. The technique improves the convergence quality, leading to better training results. We evaluate CRF-OPT on a gene prediction task on pathogenic DNA sequences, and show that it is faster and achieves better prediction accuracy than both the HMM models and the original CRF model without exponential transformation.
UR - http://www.scopus.com/inward/record.url?scp=57749178388&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:57749178388
SN - 9781577353683
T3 - Proceedings of the National Conference on Artificial Intelligence
SP - 1018
EP - 1023
BT - AAAI-08/IAAI-08 Proceedings - 23rd AAAI Conference on Artificial Intelligence and the 20th Innovative Applications of Artificial Intelligence Conference
Y2 - 13 July 2008 through 17 July 2008
ER -