A Faster Interior-Point Method for Sum-of-Squares Optimization

TL;DR

This paper introduces a faster interior-point method for sum-of-squares polynomial optimization, significantly improving computational efficiency over existing solvers by leveraging a novel dynamic data structure for Hessian inverse maintenance.

Contribution

The paper develops a new interior-point algorithm with a dynamic data structure that speeds up SOS polynomial optimization, outperforming previous semidefinite programming methods.

Findings

Algorithm runs in  O(LU^{1.87}) time

Achieves polynomial speedup over state-of-the-art SOS solvers

Extends to multivariate SOS optimization with spectral approximations

Abstract

We present a faster interior-point method for optimizing sum-of-squares (SOS) polynomials, which are a central tool in polynomial optimization and capture convex programming in the Lasserre hierarchy. Let $p = \sum_i q^2_i$ be an $n$-variate SOS polynomial of degree $2d$. Denoting by $L := \binom{n+d}{d}$ and $U := \binom{n+2d}{2d}$ the dimensions of the vector spaces in which $q_i$'s and $p$ live respectively, our algorithm runs in time $\tilde{O}(LU^{1.87})$. This is polynomially faster than state-of-art SOS and semidefinite programming solvers, which achieve runtime $\tilde{O}(L^{0.5}\min\{U^{2.37}, L^{4.24}\})$. The centerpiece of our algorithm is a dynamic data structure for maintaining the inverse of the Hessian of the SOS barrier function under the polynomial interpolant basis, which efficiently extends to multivariate SOS optimization, and requires maintaining spectral approximations to low-rank perturbations of elementwise (Hadamard) products. This is the main challenge and departure from recent IPM breakthroughs using inverse-maintenance, where low-rank updates to the slack matrix readily imply the same for the Hessian matrix.