Single-level Adversarial Data Synthesis based on Neural Tangent Kernels

Yu-Rong Zhang, Reddy Su, Sheng-Yen Chou, Shan-Hung Wu

Keywords: Adversarial, Data Synthesis, Neural Tangent Kernels

Abstract: Generative adversarial networks (GANs) have achieved impressive performance in data synthesis and have driven the development of many applications. How- ever, GANs are known to be hard to train due to their bilevel objective, which leads to the problems of convergence, mode collapse, and gradient vanishing. In this paper, we propose a new generative model called the generative adversarial NTK (GA-NTK) that has a single-level objective. The GA-NTK keeps the spirit of adversarial learning (which helps generate plausible data) while avoiding the training difficulties of GANs. This is done by modeling the discriminator as a Gaussian process with a neural tangent kernel (NTK-GP) whose training dynam- ics can be completely described by a closed-form formula. We analyze the conver- gence behavior of GA-NTK trained by gradient descent and give some sufficient conditions for convergence. We also conduct extensive experiments to study the advantages and limitations of GA-NTK and propose some techniques that make GA-NTK more practical.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Generative models

TL;DR: This paper formulates the adversarial data synthesis as a single-level optimization problem that is much easier to train than existing GANs.