Abstract:We undertake a systematic study of sketching a quadratic form: given an $n \times n$ matrix $A$, create a succinct sketch $\textbf{sk}(A)$ which can produce (without further access to $A$) a multiplicative $(1+\epsilon)$-approximation to $x^T A x$ for any desired query $x \in \mathbb{R}^n$. While a general matrix does not admit non-trivial sketches, positive semi-definite (PSD) matrices admit sketches of size $\Theta(\epsilon^{-2} n)$, via the Johnson-Lindenstrauss lemma, achieving the "for each" guarantee, namely, for each query $x$, with a constant probability the sketch succeeds. (For the stronger "for all" guarantee, where the sketch succeeds for all $x$'s simultaneously, again there are no non-trivial sketches.) We design significantly better sketches for the important subclass of graph Laplacian matrices, which we also extend to symmetric diagonally dominant matrices. A sequence of work culminating in that of Batson, Spielman, and Srivastava (SIAM Review, 2014), shows that by choosing and reweighting $O(\epsilon^{-2} n)$ edges in a graph, one achieves the "for all" guarantee. Our main results advance this front. $\bullet$ For the "for all" guarantee, we prove that Batson et al.'s bound is optimal even when we restrict to "cut queries" $x\in \{0,1\}^n$. In contrast, previous lower bounds showed the bound only for {\em spectral-sparsifiers}. $\bullet$ For the "for each" guarantee, we design a sketch of size $\tilde O(\epsilon^{-1} n)$ bits for "cut queries" $x\in \{0,1\}^n$. We prove a nearly-matching lower bound of $\Omega(\epsilon^{-1} n)$ bits. For general queries $x \in \mathbb{R}^n$, we construct sketches of size $\tilde{O}(\epsilon^{-1.6} n)$ bits.

When Lanczos Iterations Generate Symmetric Quadrature Nodes?

An analysis on stochastic Lanczos quadrature with asymmetric quadrature nodes

A Golub-Welsch version for simultaneous Gaussian quadrature

Computational aspects of simultaneous Gaussian quadrature

Very high-order symmetric positive-interior quadrature rules on triangles and tetrahedra

Quadratic transformations of Macdonald and Koornwinder polynomials

On Lebesgue Integral Quadrature

Large sparse symmetric eigenvalue problems with homogeneous linear constraints: the Lanczos process with inner–outer iterations

Error bounds for Gauss–Lobatto quadrature of analytic functions on an ellipse

Randomized matrix-free quadrature: unified and uniform bounds for stochastic Lanczos quadrature and the kernel polynomial method

Error bounds for Gauss–Jacobi quadrature of analytic functions on an ellipse

On quadrature for singular integral operators with complex symmetric quadratic forms

Network Analysis via Partial Spectral Factorization and Gauss Quadrature

Quadrature Methods for Highly Oscillatory Singular Integrals

Gaussian hypergeometric functions and cyclotomic matrices involving squares over finite fields

An analysis of Ermakov-Zolotukhin quadrature using kernels

Equivariant Gauss sum of finite quadratic forms

Gaussian quadrature rules for composite highly oscillatory integrals

On Sketching Quadratic Forms

(R,S)-(Skew) Symmetric Solutions to Matrix Equation AXB=C over Quaternions