Abstract:We study lower bounds on adaptive sensing algorithms for recovering low rank matrices using linear measurements. Given an $n \times n$ matrix $A$, a general linear measurement $S(A)$, for an $n \times n$ matrix $S$, is just the inner product of $S$ and $A$, each treated as $n^2$-dimensional vectors. By performing as few linear measurements as possible on a rank-$r$ matrix $A$, we hope to construct a matrix $\hat{A}$ that satisfies $\|A - \hat{A}\|_F^2 \le c\|A\|_F^2$, for a small constant $c$. It is commonly assumed that when measuring $A$ with $S$, the response is corrupted with an independent Gaussian random variable of mean $0$ and variance $\sigma^2$. Candés and Plan study non-adaptive algorithms for low rank matrix recovery using random linear measurements. At a certain noise level, it is known that their non-adaptive algorithms need to perform $\Omega(n^2)$ measurements, which amounts to reading the entire matrix. An important question is whether adaptivity helps in decreasing the overall number of measurements. We show that any adaptive algorithm that uses $k$ linear measurements in each round and outputs an approximation to the underlying matrix with probability $\ge 9/10$ must run for $t = \Omega(\log(n^2/k)/\log\log n)$ rounds showing that any adaptive algorithm which uses $n^{2-\beta}$ linear measurements in each round must run for $\Omega(\log n/\log\log n)$ rounds to compute a reconstruction with probability $\ge 9/10$. Hence any adaptive algorithm that has $o(\log n/\log\log n)$ rounds must use an overall $\Omega(n^2)$ linear measurements. Our techniques also readily extend to obtain lower bounds on adaptive algorithms for tensor recovery and obtain measurement-vs-rounds trade-off for many sensing problems in numerical linear algebra, such as spectral norm low rank approximation, Frobenius norm low rank approximation, singular vector approximation, and more.

Low Rank Matrix Recovery with Adversarial Sparse Noise

CONVERGENCE AND STABILITY OF ITERATIVELY REWEIGHTED LEAST SQUARES FOR LOW-RANK MATRIXRECOVERY

Sparse Recovery with Coherent Tight Frame via Analysis Dantzig Selector and Analysis LASSO

I T ] 8 S ep 2 01 3 Sparse Recovery with Coherent Tight Frames via Analysis Dantzig Selector and Analysis LASSO ∗

On the Schatten P-Quasi-norm Minimization for Low-Rank Matrix Recovery

Convergence of Projected Subgradient Method with Sparse or Low-Rank Constraints

Analysis Non-Sparse Recovery for Relaxed ALASSO

Sparsity tracking for low rank matrix recovery from noise

Matrix recovery from nonconvex regularized least absolute deviations

Robust Sensing of Low-Rank Matrices with Non-Orthogonal Sparse Decomposition

Lower Bounds on Adaptive Sensing for Matrix Recovery

Robust Matrix Completion with Heavy-tailed Noise

Matrix Completion with Sparse Noisy Rows

Geometric Analysis of Noisy Low-Rank Matrix Recovery in the Exact Parametrized and the Overparametrized Regimes

Sharp Restricted Isometry Property Bounds for Low-rank Matrix Recovery Problems with Corrupted Measurements

Leveraged Matrix Completion with Noise

Convergence and stability of iteratively reweighted least squares for low-rank matrix recovery

Guaranteed matrix recovery using weighted nuclear norm plus weighted total variation minimization

Sharp Support Recovery from Noisy Random Measurements by L1 minimization

A Validation Approach to Over-parameterized Matrix and Image Recovery

Can Learning Be Explained By Local Optimality In Low-rank Matrix Recovery?