Information-Theoretic Limits of Matrix Completion

TL;DR

This paper develops an information-theoretic framework for matrix completion that applies to general matrices with low description complexity, establishing measurement bounds for successful recovery under broad conditions.

Contribution

It introduces a novel information-theoretic approach to matrix completion that extends beyond low-rank matrices to those with low description complexity, using rank-one measurements.

Findings

Recovery with high probability from fewer measurements than matrix dimensions.

Rank-one measurements are effective and computationally efficient.

Exact bounds for low-rank matrix recovery are established.

Abstract

We propose an information-theoretic framework for matrix completion. The theory goes beyond the low-rank structure and applies to general matrices of "low description complexity". Specifically, we consider $m\times n$ random matrices $\mathbf{X}$ of arbitrary distribution (continuous, discrete, discrete-continuous mixture, or even singular). With $\mathcal{S}$ an $\varepsilon$-support set of $\mathbf{X}$, i.e., $\mathrm{P}[\mathbf{X}\in\mathcal{S}]\geq 1-\varepsilon$, and $\underline{\mathrm{dim}}_\mathrm{B}(\mathcal{S})$ denoting the lower Minkowski dimension of $\mathcal{S}$, we show that $k> \underline{\mathrm{dim}}_\mathrm{B}(\mathcal{S})$ trace inner product measurements with measurement matrices $A_i$, suffice to recover $\mathbf{X}$ with probability of error at most $\varepsilon$. The result holds for Lebesgue a.a. $A_i$ and does not need incoherence between the $A_i$ and the unknown matrix $\mathbf{X}$. We furthermore show that $k> \underline{\mathrm{dim}}_\mathrm{B}(\mathcal{S})$ measurements also suffice to recover the unknown matrix $\mathbf{X}$ from measurements taken with rank-one $A_i$, again this applies to a.a. rank-one $A_i$. Rank-one measurement matrices are attractive as they require less storage space than general measurement matrices and can be applied faster. Particularizing our results to the recovery of low-rank matrices, we find that $k>(m+n-r)r$ measurements are sufficient to recover matrices of rank at most $r$. Finally, we construct a class of rank-$r$ matrices that can be recovered with arbitrarily small probability of error from $k<(m+n-r)r$ measurements.