NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:7987
Title:Outlier Detection and Robust PCA Using a Convex Measure of Innovation


		
This paper presents a new convex approach to robust PCA that is different from previously proposed convex approaches, and is shown to have good theoretical performance guarantees under several different outlier distributions. The paper seems careful and a new contribution (inspire of the method, like other convex approaches, being somewhat high-complexity). the reviewers are also in agreement that this should be accepted.