Loading...
Please wait, while we are loading the content...
Similar Documents
L G ] 2 3 O ct 2 01 8 Finding Mixed Nash Equilibria of Generative Adversarial Networks
| Content Provider | Semantic Scholar |
|---|---|
| Author | Hsieh, Ya-Ping Liu, Cong Shan |
| Copyright Year | 2018 |
| Abstract | We reconsider the training objective of Generative Adversarial Networks (GANs) from the mixed Nash Equilibria (NE) perspective. Inspired by the classical prox methods, we develop a novel algorithmic framework for GANs via an infinite-dimensional two-player game and prove rigorous convergence rates to the mixed NE, resolving the longstanding problem that no provably convergent algorithm exists for general GANs. We then propose a principled procedure to reduce our novel prox methods to simple sampling routines, leading to practically efficient algorithms. Finally, we provide experimental evidence that our approach outperforms methods that seek pure strategy equilibria, such as SGD, Adam, and RMSProp, both in speed and quality. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://export.arxiv.org/pdf/1811.02002 |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
