Distributing mass under a pointwise bound and an application to weighted polynomial approximation

TL;DR

This paper develops a duality framework for optimal mass distribution under pointwise and total mass constraints, applying it to weighted polynomial approximation problems and confirming a longstanding conjecture.

Contribution

It introduces a duality theorem linking mass distribution optimization to semi-covers by cubes, advancing understanding in weighted polynomial approximation.

Findings

Proves a duality theorem for mass distribution problems.

Solves the splitting problem in weighted polynomial approximation.

Confirms a conjecture by Kriete and MacCluer.

Abstract

Inspired by applications in weighted polynomial approximation problems, we study an optimal mass distribution problem. Given a gauge function $h$ and a positive "roof" function $R$ compactly supported in $\mathbb{R}^n$, we are interested in estimating the supremum of the $L^1$-norms of non-negative functions $f$ satisfying the pointwise bound $f \leq R$ and the mass distribution bound $\int_c f \, dV_n \leq h(V_n(c))$, where $c$ is a cube and $V_n$ is the volume measure. We prove a duality theorem which states that the optimal value in this maximization problem is the minimum among certain quantities associated with semi-covers by cubes of the support of $R$. We use our theorem to solve the so-called "splitting problem" in the theory of polynomial approximations in the plane. As a result, we confirm an old conjecture of Kriete and MacCluer regarding an extension of Khrushchev's original splitting theorem to the weighted context, and explain the mechanics of an example in the research problem book by Havin, Khrushchev and Nikolskii.