Dual Supervised Learning

Yingce Xia, Tao Qin, Wei Chen, Jiang Bian, Nenghai Yu, Tie-Yan Liu
Proceedings of the 34th International Conference on Machine Learning, PMLR 70:3789-3798, 2017.

Abstract

Many supervised learning tasks are emerged in dual forms, e.g., English-to-French translation vs. French-to-English translation, speech recognition vs. text to speech, and image classification vs. image generation. Two dual tasks have intrinsic connections with each other due to the probabilistic correlation between their models. This connection is, however, not effectively utilized today, since people usually train the models of two dual tasks separately and independently. In this work, we propose training the models of two dual tasks simultaneously, and explicitly exploiting the probabilistic correlation between them to regularize the training process. For ease of reference, we call the proposed approach dual supervised learning. We demonstrate that dual supervised learning can improve the practical performances of both tasks, for various applications including machine translation, image processing, and sentiment analysis.

Cite this Paper

BibTeX
@InProceedings{pmlr-v70-xia17a, title = {Dual Supervised Learning}, author = {Yingce Xia and Tao Qin and Wei Chen and Jiang Bian and Nenghai Yu and Tie-Yan Liu}, booktitle = {Proceedings of the 34th International Conference on Machine Learning}, pages = {3789--3798}, year = {2017}, editor = {Precup, Doina and Teh, Yee Whye}, volume = {70}, series = {Proceedings of Machine Learning Research}, month = {06--11 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v70/xia17a/xia17a.pdf}, url = {https://proceedings.mlr.press/v70/xia17a.html}, abstract = {Many supervised learning tasks are emerged in dual forms, e.g., English-to-French translation vs. French-to-English translation, speech recognition vs. text to speech, and image classification vs. image generation. Two dual tasks have intrinsic connections with each other due to the probabilistic correlation between their models. This connection is, however, not effectively utilized today, since people usually train the models of two dual tasks separately and independently. In this work, we propose training the models of two dual tasks simultaneously, and explicitly exploiting the probabilistic correlation between them to regularize the training process. For ease of reference, we call the proposed approach dual supervised learning. We demonstrate that dual supervised learning can improve the practical performances of both tasks, for various applications including machine translation, image processing, and sentiment analysis.} }
Endnote
%0 Conference Paper %T Dual Supervised Learning %A Yingce Xia %A Tao Qin %A Wei Chen %A Jiang Bian %A Nenghai Yu %A Tie-Yan Liu %B Proceedings of the 34th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2017 %E Doina Precup %E Yee Whye Teh %F pmlr-v70-xia17a %I PMLR %P 3789--3798 %U https://proceedings.mlr.press/v70/xia17a.html %V 70 %X Many supervised learning tasks are emerged in dual forms, e.g., English-to-French translation vs. French-to-English translation, speech recognition vs. text to speech, and image classification vs. image generation. Two dual tasks have intrinsic connections with each other due to the probabilistic correlation between their models. This connection is, however, not effectively utilized today, since people usually train the models of two dual tasks separately and independently. In this work, we propose training the models of two dual tasks simultaneously, and explicitly exploiting the probabilistic correlation between them to regularize the training process. For ease of reference, we call the proposed approach dual supervised learning. We demonstrate that dual supervised learning can improve the practical performances of both tasks, for various applications including machine translation, image processing, and sentiment analysis.
APA
Xia, Y., Qin, T., Chen, W., Bian, J., Yu, N. & Liu, T.. (2017). Dual Supervised Learning. Proceedings of the 34th International Conference on Machine Learning, in Proceedings of Machine Learning Research 70:3789-3798 Available from https://proceedings.mlr.press/v70/xia17a.html.

Related Material