NeurIPS 2020

Fourier Sparse Leverage Scores and Approximate Kernel Learning

Meta Review

Four knowledgeable reviewers recommend to accept the paper based on the progress it has made toward understanding and improving the accuracy of kernel approximations made using leverage scores. On this basis the paper is accepted. We ask that the authors implement the edits proposed in their rebuttal: to clarify the justification for the equation after line 518, and to expand on the discussion of Theorem 3.