Tight Analysis of Extra-gradient and Optimistic Gradient Methods For Nonconvex Minimax Problems.
Pouria MahdaviniaYuyang DengHaochuan LiMehrdad MahdaviPublished in: NeurIPS (2022)
Keyphrases
- difficult problems
- steepest ascent
- methods can be applied
- statistical analysis
- machine learning methods
- analytical methods
- computational cost
- optimization problems
- image analysis
- gradient method
- neural network
- problems in computer vision
- numerical methods
- significant improvement
- lower bound
- np complete
- practical problems
- data mining techniques
- edge detection
- convergence analysis
- objective function
- data sets
- image restoration and reconstruction