Non-Ergodic Alternating Proximal Augmented Lagrangian Algorithms with Optimal Rates

Quoc Tran Dinh

Neural Information Processing Systems 

Problem (1) issufficiently machine coverscon compressi OurapprOurapproach function incorporate 17] or [25]) into andy. T known accelerated different x and y. non-er Related Our Lagrangian-type aremost of Po method alternating and ADMM Rachford' 8,15] or practice, using e.g., [8], itsO Thedualproblemof (1) is d?:= min

Similar Docs  Excel Report  more

TitleSimilaritySource
None found