Estimating the Spectral Density of Large Implicit Matrices

by   Ryan P. Adams, et al.

Many important problems are characterized by the eigenvalues of a large matrix. For example, the difficulty of many optimization problems, such as those arising from the fitting of large models in statistics and machine learning, can be investigated via the spectrum of the Hessian of the empirical loss function. Network data can be understood via the eigenstructure of a graph Laplacian matrix using spectral graph theory. Quantum simulations and other many-body problems are often characterized via the eigenvalues of the solution space, as are various dynamic systems. However, naive eigenvalue estimation is computationally expensive even when the matrix can be represented; in many of these situations the matrix is so large as to only be available implicitly via products with vectors. Even worse, one may only have noisy estimates of such matrix vector products. In this work, we combine several different techniques for randomized estimation and show that it is possible to construct unbiased estimators to answer a broad class of questions about the spectra of such implicit matrices, even in the presence of noise. We validate these methods on large-scale problems in which graph theory and random matrix theory provide ground truth.


page 1

page 2

page 3

page 4


Approximating matrix eigenvalues by randomized subspace iteration

Traditional numerical methods for calculating matrix eigenvalues are pro...

Approximate Graph Spectral Decomposition with the Variational Quantum Eigensolver

Spectral graph theory is a branch of mathematics that studies the relati...

Shrinkage Estimation of Functions of Large Noisy Symmetric Matrices

We study the problem of estimating functions of a large symmetric matrix...

Spectral density of random graphs: convergence properties and application in model fitting

Random graph models are used to describe the complex structure of real-w...

NFFT meets Krylov methods: Fast matrix-vector products for the graph Laplacian of fully connected networks

The graph Laplacian is a standard tool in data science, machine learning...

The Wigner's Semicircle Law of Weighted Random Networks

The spectral graph theory provides an algebraical approach to investigat...

Dynamic Trace Estimation

We study a dynamic version of the implicit trace estimation problem. Giv...

Please sign up or login with your details

Forgot password? Click here to reset