NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:3136
Title:Generalized Block-Diagonal Structure Pursuit: Learning Soft Latent Task Assignment against Negative Transfer


		
The authors present a novel framework to address the issue of negative transfer when conducting multi-task learning. All reviewers agree that the setting is useful and that the algorithm is well motivated and has demonstrated value. The main concern is around clarity of some parts of the paper and the ease of interpreting some of the theoretical results. The author feedback has addressed most of the questions. I strongly expect the final version is expected to include these clarifications as well as much additional polishing as possible.