ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching
Abstract
We investigate the non-identifiability issues associated with bidirectional adversarial training for joint distribution matching. Within a framework of conditional entropy, we propose both adversarial and non-adversarial approaches to learn desirable matched joint distributions for unsupervised and supervised tasks. We unify a broad family of adversarial models as joint distribution matching problems. Our approach stabilizes learning of unsupervised bidirectional adversarial learning methods. Further, we introduce an extension for semi-supervised learning tasks. Theoretical results are validated in synthetic data and real-world applications.
Cite
Text
Li et al. "ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching." Neural Information Processing Systems, 2017.Markdown
[Li et al. "ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching." Neural Information Processing Systems, 2017.](https://mlanthology.org/neurips/2017/li2017neurips-alice/)BibTeX
@inproceedings{li2017neurips-alice,
title = {{ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching}},
author = {Li, Chunyuan and Liu, Hao and Chen, Changyou and Pu, Yuchen and Chen, Liqun and Henao, Ricardo and Carin, Lawrence},
booktitle = {Neural Information Processing Systems},
year = {2017},
pages = {5495-5503},
url = {https://mlanthology.org/neurips/2017/li2017neurips-alice/}
}