Learning Greedy Policies for the Easy-First Framework

Jun Xie, Chao Ma, Janardhan Rao Doppa, Prashanth Mannem, Xiaoli Z. Fern, Thomas G. Dietterich, Prasad Tadepalli

AAAI 2015 pp. 2339-2345

doi:10.1609/AAAI.V29I1.9509 /aaai/2015/xie2015aaai-learning/

Abstract

Easy-first, a search-based structured prediction approach, has been applied to many NLP tasks including dependency parsing and coreference resolution. This approach employs a learned greedy policy (action scoring function) to make easy decisions first, which constrains the remaining decisions and makes them easier. We formulate greedy policy learning in the Easy-first approach as a novel non-convex optimization problem and solve it via an efficient Majorization Minimizatoin (MM) algorithm. Results on within-document coreference and cross-document joint entity and event coreference tasks demonstrate that the proposed approach achieves statistically significant performance improvement over existing training regimes for Easy-first and is less susceptible to overfitting.

PDF AAAI Semantic Scholar

Cite

Text

Xie et al. "Learning Greedy Policies for the Easy-First Framework." AAAI Conference on Artificial Intelligence, 2015. doi:10.1609/AAAI.V29I1.9509

Markdown

[Xie et al. "Learning Greedy Policies for the Easy-First Framework." AAAI Conference on Artificial Intelligence, 2015.](https://mlanthology.org/aaai/2015/xie2015aaai-learning/) doi:10.1609/AAAI.V29I1.9509

BibTeX

@inproceedings{xie2015aaai-learning,
  title     = {{Learning Greedy Policies for the Easy-First Framework}},
  author    = {Xie, Jun and Ma, Chao and Doppa, Janardhan Rao and Mannem, Prashanth and Fern, Xiaoli Z. and Dietterich, Thomas G. and Tadepalli, Prasad},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2015},
  pages     = {2339-2345},
  doi       = {10.1609/AAAI.V29I1.9509},
  url       = {https://mlanthology.org/aaai/2015/xie2015aaai-learning/}
}