NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
The authors study the trajectories of intermediate variables in ADMM, and use this to propose a refinement to inertial ADMM. The paper is full of interesting ideas and the reviews were positive. One reviewer expressed concerns about the paper's readability and presentation; I would like the authors to address this to the extent they can. [This meta-review was reviewed and revised by the Program Chairs]