In , one is given and binary mask matrix . The goal is to find a rank- matrix for which: cost(L) = \sum_{i=1}^{n} \sum_{j = 1}^{n} W_{i,j} \cdot (A_{i,j} - L_{i,j} )^2 \leq OPT + \epsilon \|A\|_F^2 , where and is a given error parameter. Depending on the choice of , this problem captures factor analysis, low-rank plus diagonal decomposition, robust PCA, low-rank matrix completion, low-rank plus block matrix approximation, and many problems. Many of these problems are NP-hard, and while some algorithms with provable guarantees are known, they either 1) run in time or 2) make strong assumptions, e.g., that is incoherent or that is random. In this work, we show that a common polynomial time heuristic, which simply sets to where is , and then finds a standard low-rank approximation, yields bicriteria approximation guarantees for this problem. In particular, for rank depending on the of , the heuristic outputs rank- with cost. This partition number is in turn bounded by the of , when interpreted as a two-player communication matrix. For many important examples of masked low-rank approximation, including all those listed above, this result yields bicriteria approximation guarantees with . Further, we show that different models of communication yield algorithms for natural variants of masked low-rank approximation. For example, multi-player number-in-hand communication complexity connects to masked tensor decomposition and non-deterministic communication complexity to masked Boolean low-rank factorization.
View on arXiv