On U-Statistics and Compressed Sensing II: Non-Asymptotic Worst-Case Analysis

TL;DR

This paper introduces a novel non-asymptotic worst-case analysis for compressed sensing using U-statistics, establishing tight union bounds for restricted isometries and mutual coherence, which enhances theoretical understanding of measurement bounds.

Contribution

It presents the first theoretical U-statistical results for worst-case analysis in compressed sensing, demonstrating tight union bounds for key parameters like restricted isometry constants and mutual coherence.

Findings

Restricted isometry constants have tight union bounds when measurements are proportional to k (1 + log(n/k)).

Mutual coherence estimates also have very tight union bounds in Gaussian and Bernoulli cases.

The results suggest potential improvements in measurement bounds and theoretical guarantees in compressed sensing.

Abstract

In another related work, U-statistics were used for non-asymptotic "average-case" analysis of random compressed sensing matrices. In this companion paper the same analytical tool is adopted differently - here we perform non-asymptotic "worst-case" analysis. Simple union bounds are a natural choice for "worst-case" analyses, however their tightness is an issue (and questioned in previous works). Here we focus on a theoretical U-statistical result, which potentially allows us to prove that these union bounds are tight. To our knowledge, this kind of (powerful) result is completely new in the context of CS. This general result applies to a wide variety of parameters, and is related to (Stein-Chen) Poisson approximation. In this paper, we consider i) restricted isometries, and ii) mutual coherence. For the bounded case, we show that k-th order restricted isometry constants have tight union bounds, when the measurements m = \mathcal{O}(k (1 + \log(n/k))). Here we require the restricted isometries to grow linearly in k, however we conjecture that this result can be improved to allow them to be fixed. Also, we show that mutual coherence (with the standard estimate \sqrt{(4\log n)/m}) have very tight union bounds. For coherence, the normalization complicates general discussion, and we consider only Gaussian and Bernoulli cases here.