NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:7077
Title:Kernel quadrature with DPPs


		
The paper proposes a kernel quadrature method with determinantal point processes, and provide some theoretical and experimental results. We think that the paper shows promising experimental results, but the obtained theoretical results may not be appropriately compared with the recent works on Bayesian quadrature. We recommend acceptance for this submission assuming that the final camera ready reflect the following points, which have been discussed in the Authors' Feedback. - Include references on Bayesian quadrature to reflect the recent advances, and discuss the relation with this work. - Discuss the known optimal rate for Bayesian quadrature including the one with uniform grids, and compare it with the current results. - Include the experimental results of Figure A and relevant discussions in the Authors' Feedback.