Abstract:For a large Hermitian matrix $A\in \mathbb{C}^{N\times N}$, it is often the case that the only affordable operation is matrix-vector multiplication. In such case, randomized method is a powerful way to estimate the spectral density (or density of states) of $A$. However, randomized methods developed so far for estimating spectral densities only extract information from different random vectors independently, and the accuracy is therefore inherently limited to $\mathcal{O}(1/\sqrt{N_{v}})$ where $N_{v}$ is the number of random vectors. In this paper we demonstrate that the "$\mathcal{O}(1/\sqrt{N_{v}})$ barrier" can be overcome by taking advantage of the correlated information of random vectors when properly filtered by polynomials of $A$. Our method uses the fact that the estimation of the spectral density essentially requires the computation of the trace of a series of matrix functions that are numerically low rank. By repeatedly applying $A$ to the same set of random vectors and taking different linear combination of the results, we can sweep through the entire spectrum of $A$ by building such low rank decomposition at different parts of the spectrum. Under some assumptions, we demonstrate that a robust and efficient implementation of such spectrum sweeping method can compute the spectral density accurately with $\mathcal{O}(N^2)$ computational cost and $\mathcal{O}(N)$ memory cost. Numerical results indicate that the new method can significantly outperform existing randomized methods in terms of accuracy. As an application, we demonstrate a way to accurately compute a trace of a smooth matrix function, by carefully balancing the smoothness of the integrand and the regularized density of states using a deconvolution procedure.

On Spectral Portraits of Incidence Matrices of Nearest Neighbor Graphs

Spectral Inference Methods on Sparse Graphs: Theory and Applications

Spectral Theory of Sparse Non-Hermitian Random Matrices

Spectral radii of sparse random matrices

On Random Matrices Arising in Deep Neural Networks. Gaussian Case

On Random Matrices Arising in Deep Neural Networks: General I.I.D. Case

Network Analysis via Partial Spectral Factorization and Gauss Quadrature

Isospectral Reductions of Non-negative Matrices

On the rank, Kernel, and core of sparse random graphs

Sublinear Time Spectral Density Estimation

A Spectral Graph Uncertainty Principle

On randomized estimators of the Hafnian of a nonnegative matrix

Varentropy Estimation via Nearest Neighbor Graphs

Planar Novikov-Shubin invariant for adjacency matrices of structured directed dense random graphs

Randomized estimation of spectral densities of large matrices made accurate

Optimal construction of k-nearest neighbor graphs for identifying noisy clusters

The Spectral Edge of Constant Degree Erdős-Rényi Graphs

Lecture Notes on Spectral Independence and Bases of a Matroid: Local-to-Global and Trickle-Down from a Markov Chain Perspective

Rank-one matrix estimation: analysis of algorithmic and information theoretic limits by the spatial coupling method

Universality for random matrices with an edge spectrum singularity

The Exact Rank of Sparse Random Graphs