Learning generative models across incomparable spaces
Name
1905.05461.pdf
Description
Accepted version
Size
8.85 MB
Format
Unknown
Checksum (MD5)
654abc698c668c449e08a27f812b3a3e
Author(s)
Bunne, Charlotte
Alvarez Melis, David
Krause, Andreas
Jegelka, Stefanie Sabrina
Date Issued
2019
Journal
36th International Conference on Machine Learning, ICML 2019
Version
Author's final manuscript
Abstract
© 36th International Conference on Machine Learning, ICML 2019. All rights reserved. Generative Adversarial Networks have shown remarkable success in learning a distribution that faithfully recovers a reference distribution in its entirety. However, in some cases, we may want to only learn some aspects (e.g., cluster or manifold structure), while modifying others (e.g., style, orientation or dimension). In this work, we propose an approach to learn generative models across such incomparable spaces, and demonstrate how to steer the learned distribution towards target properties. A key component of our model is the Gromov-Wasserstein distance, a notion of discrepancy that compares distributions relationally rather than absolutely. While this framework subsumes current generative models in identically reproducing distributions, its inherent flexibility allows application to tasks in manifold learning, relational learning and cross-domain learning.
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
http://proceedings.mlr.press/v97/bunne19a
