Reviews: Subset Selection under Noise

This paper considers the problem of maximizing a monotone set function subject to a cardinality constraint. The authors consider a novel combination of functions with both bounded submodularity ratio and additive noise. These setting have been considered separately before, but a joint analysis leads to a novel algorithm PONSS. This has improved theoretical guarantees and experimental performance when compared to previous noise-agnostic greedy algorithms. The paper flows well and is generally a pleasure to read.