Training Generative Adversarial Networks via Primal-Dual subgradient Methods: a Lagrangian Perspective on GaN.

Xu Chen Jiang Wang Hao Ge

Published in: ICLR (Poster) (2018)

Keyphrases

primal dual
optimization methods
convex optimization
variational inequalities
convergence analysis
convex optimization problems
upper bound
semidefinite programming
interior point methods