Sharma, A and Parekh, Z and Talukdar, P (2017) Speeding up reinforcement learning-based information extraction training using asynchronous methods. In: 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, 9 - 11 September 2017, Copenhagen, pp. 2658-2663.
|
PDF
spe_Rei_lea_Inf_ext_2017.pdf - Published Version Download (670kB) | Preview |
Abstract
RLIE-DQN is a recently proposed Reinforcement Learning-based Information Extraction (IE) technique which is able to incorporate external evidence during the extraction process. RLIE-DQN trains a single agent sequentially, training on one instance at a time. This results in significant training slowdown which is undesirable. We leverage recent advances in parallel RL training using asynchronous methods and propose RLIE-A3C. RLIEA3C trains multiple agents in parallel and is able to achieve upto 6x training speedup over RLIE-DQN, while suffering no loss in average accuracy.
| Item Type: | Conference Paper |
|---|---|
| Publication: | EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings |
| Publisher: | Association for Computational Linguistics (ACL) |
| Additional Information: | The copyright for this article belongs to Association for Computational Linguistics (ACL) |
| Keywords: | Information retrieval; Machine learning; Multi agent systems; Natural language processing systems, Asynchronous methods; Extraction process; Information extraction techniques; Multiple agents; Single-agent, Reinforcement learning |
| Department/Centre: | Division of Interdisciplinary Sciences > Computational and Data Sciences |
| Date Deposited: | 19 Jul 2022 11:53 |
| Last Modified: | 19 Jul 2022 11:53 |
| URI: | https://eprints.iisc.ac.in/id/eprint/74905 |
Actions (login required)
 |
View Item |
