Multi-Source Domain Adaptation with Mixture of Experts
Name
D18-1498.pdf
Description
Published version
Size
882.58 KB
Format
Adobe PDF
Checksum (MD5)
183aa6abf07fd7895be83da7d72fac60
Author(s)
Guo, Jiang
Shah, Darsh
Barzilay, Regina
Date Issued
October 2018
Journal
Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, EMNLP 2018
Publisher
Association for Computational Linguistics (ACL)
Citation
Guo, Jiang, Shah, Darsh and Barzilay, Regina. 2018. "Multi-Source Domain Adaptation with Mixture of Experts." Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, EMNLP 2018.
Version
Final published version
Abstract
© 2018 Association for Computational Linguistics We propose a mixture-of-experts approach for unsupervised domain adaptation from multiple sources. The key idea is to explicitly capture the relationship between a target example and different source domains. This relationship, expressed by a point-to-set metric, determines how to combine predictors trained on various domains. The metric is learned in an unsupervised fashion using meta-training. Experimental results on sentiment analysis and part-of-speech tagging demonstrate that our approach consistently outperforms multiple baselines and can robustly handle negative transfer.1
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Terms of Use
Creative Commons Attribution 4.0 International license
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.18653/V1/D18-1498
