NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:2824
Title:Optimistic Regret Minimization for Extensive-Form Games via Dilated Distance-Generating Functions


		
This paper proposes the use of "dilated" distance generating functions to write the associated mirror descent algorithm as a local regret minimization procedure on the decision nodes in two-person zero-sum extensive form games. After the discussion phase, the reviewers and myself did not have any major concerns for the paper and I am happy to recommend acceptance (modulo the reviewers' specific suggestions for points to be revised in the camera-ready).