Asymptotics of the Sketched Pseudoinverse

TL;DR

This paper uses random matrix theory to analyze the asymptotic behavior of the sketched pseudoinverse of positive semidefinite matrices, providing new insights into its eigenvalues and applications in regression and iterative methods.

Contribution

It extends previous asymptotic equivalence results to real-valued regularization, including negative regularization, and characterizes second-order equivalence and eigenvalue behavior.

Findings

Asymptotic first-order equivalence of sketched pseudoinverse to resolvent evaluation.

Precise characterization of the smallest nonzero eigenvalue of the sketched matrix.

Application of results to sketch-and-project and sketched ridge regression.

Abstract

We take a random matrix theory approach to random sketching and show an asymptotic first-order equivalence of the regularized sketched pseudoinverse of a positive semidefinite matrix to a certain evaluation of the resolvent of the same matrix. We focus on real-valued regularization and extend previous results on an asymptotic equivalence of random matrices to the real setting, providing a precise characterization of the equivalence even under negative regularization, including a precise characterization of the smallest nonzero eigenvalue of the sketched matrix, which may be of independent interest. We then further characterize the second-order equivalence of the sketched pseudoinverse. We also apply our results to the analysis of the sketch-and-project method and to sketched ridge regression. Lastly, we prove that these results generalize to asymptotically free sketching matrices, obtaining the resulting equivalence for orthogonal sketching matrices and comparing our results to several common sketches used in practice.