Efficient deterministic approximate counting for low-degree polynomial threshold functions

TL;DR

This paper presents a deterministic polynomial-time algorithm for approximately counting satisfying assignments of low-degree polynomial threshold functions, improving efficiency over previous methods by leveraging new probabilistic and polynomial decomposition techniques.

Contribution

The work introduces a novel multivariate CLT using Malliavin calculus and Stein's Method, and a polynomial decomposition approach, enabling a fixed-polynomial time algorithm for approximate counting.

Findings

Achieved a deterministic algorithm with fixed polynomial runtime for approximate counting.

Developed a new multivariate CLT showing Gaussian approximation for polynomials with small eigenvalues.

Devised a polynomial decomposition method to facilitate Gaussian-space analysis.

Abstract

We give a deterministic algorithm for approximately counting satisfying assignments of a degree-$d$ polynomial threshold function (PTF). Given a degree-$d$ input polynomial $p(x_1,\dots,x_n)$ over $R^n$ and a parameter $\epsilon> 0$, our algorithm approximates $\Pr_{x \sim \{-1,1\}^n}[p(x) \geq 0]$ to within an additive $\pm \epsilon$ in time $O_{d,\epsilon}(1)\cdot \mathop{poly}(n^d)$. (Any sort of efficient multiplicative approximation is impossible even for randomized algorithms assuming $NP\not=RP$.) Note that the running time of our algorithm (as a function of $n^d$, the number of coefficients of a degree-$d$ PTF) is a \emph{fixed} polynomial. The fastest previous algorithm for this problem (due to Kane), based on constructions of unconditional pseudorandom generators for degree-$d$ PTFs, runs in time $n^{O_{d,c}(1) \cdot \epsilon^{-c}}$ for all $c > 0$. The key novel contributions of this work are: A new multivariate central limit theorem, proved using tools from Malliavin calculus and Stein's Method. This new CLT shows that any collection of Gaussian polynomials with small eigenvalues must have a joint distribution which is very close to a multidimensional Gaussian distribution. A new decomposition of low-degree multilinear polynomials over Gaussian inputs. Roughly speaking we show that (up to some small error) any such polynomial can be decomposed into a bounded number of multilinear polynomials all of which have extremely small eigenvalues. We use these new ingredients to give a deterministic algorithm for a Gaussian-space version of the approximate counting problem, and then employ standard techniques for working with low-degree PTFs (invariance principles and regularity lemmas) to reduce the original approximate counting problem over the Boolean hypercube to the Gaussian version.