The truncated K-moment problem for closure of open sets

DOI10.1016/J.JFA.2012.09.001MaRDI QIDQ694766zbMATH OpenOpenAlexFDO

Authors Grigoriy Blekherman, Jean B. Lasserre

Publication date 13 December 2012

Published in Journal of Functional Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1108.0627

zbMATH Keywords

Legendre-Fenchel transform maximum entropy moment problem truncated moment problem

Mathematics Subject Classification ID

Moment problems (44A60) Linear operator methods in interpolation, moment and extension problems (47A57)

Abstract: We solve the truncated K-moment problem when

K s u b s e t e q R^{n}

is the closure of a, not necessarily bounded, open set (which includes the important cases

K = R^{n}

and

K = R_{+}^{n}

). That is, we completely characterize the interior of the convex cone of finite sequences that have a representing measure on

K s u b s e t e q R^{n}

. It is in fact the domain of the Legendre-Fenchel transform associated with a certain convex function. And so in this context, detecting whether a sequence is in the interior of this cone reduces to solving a finite-dimensional convex optimization problem. This latter problem is related to maximum entropy methods for approximating an unknown density from knowing only finitely many of its moments. Interestingly, the proposed approach is essentially geometric and of independent interest, as it also addresses the abstract problem of characterizing the interior of a convex cone C which is the conical hull of a set continuously parametrized by a compact set

M s u b s e t R^{n}

or

M s u b s e t e q S^{n - 1}

, where

M

is the closure of an open subset of

R^{n}

(resp.

S^{n - 1}

). As a by-product we also obtain a barrier function for the cone C.

Review text

Let \(K\) be a closed set of \({\mathbb R}^n\) and let \(y=(y_\alpha)\) be a finite sequence \(\alpha\in {\mathbb N}^n_{2d}:=\{\alpha\in {\mathbb N}^n:\sum_i\alpha_i\leq 2d\}\). The truncated \(K\)-moment problem is to find a Borel measure on \(K\) such that \(y_\alpha=\int_K x^\alpha \,d\sigma(x)\) \((\alpha\in {\mathbb N}^n_{2d})\). A sequence \(y=(y_\alpha)\) which admits such a representation will be called \(K\)-moment sequence. In the case when \(n=1\) and \(K\) coincides with \((-\infty,\infty)\), \([0,\infty)\) or \([0,1]\), the corresponding problem is called the Hamburger, Stieltjes or Hausdorff moment problem, respectively. In the present paper, the authors find a new characterization of the set of \(K\)-moment sequences in terms of some optimization problem. Namely, if \(K\) is a compact set, \(\mu\) is a finite measure on \(K\) absolutely continuous with respect to the Lebesgue measure, then a sequence \(y=(y_\alpha)\) belongs to the interior of the set of all \(K\)-moment sequences if and only if \(y=(y_\alpha)\) belongs to the domain of the Legendre-Fenchel transform of the function \(f:p\mapsto f(p)=\int_Ke^{p(x)}\,d\mu(x)\) defined on the set of all polynomials \(p\) of formal degree \(2d\). Some generalizations of this result to the case of unbounded \(K\) are also presented.

Recommendations

Cites work

Cited in

(8)

This page was built for publication: The truncated K-moment problem for closure of open sets

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q694766)