Sharp finiteness principles for Lipschitz selections (Q1632236)

From MaRDI portal
Revision as of 02:39, 18 August 2024 by Daniel (talk | contribs) (‎Created claim: Wikidata QID (P12): Q129237682, #quickstatements; #temporary_batch_1723941302328)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
scientific article
Language Label Description Also known as
English
Sharp finiteness principles for Lipschitz selections
scientific article

    Statements

    Sharp finiteness principles for Lipschitz selections (English)
    0 references
    0 references
    0 references
    13 December 2018
    0 references
    The authors prove the finiteness theorem for Lipschitz selection problems, conjectured by \textit{Y. Brudnyi} and \textit{P. Shvartsman} [Int. Math. Res. Not. 1994, No. 3, 129--139 (1994; Zbl 0845.57022)]. This has previously been established by \textit{C. Fefferman} et al. [Geom. Funct. Anal. 26, No. 2, 422--477 (2016; Zbl 1353.58004)] and Shvartsman in a sequence of papers in some particular cases. The result is stated in full generality. Let \(\mathcal K_m(Y)\) stand for the family of nonempty compact convex subsets \(K\subset Y\) contained in an affine subspace of dimension at most \(m\). Given \(m\geq 1\) and \(\lambda>0\), a Banach space \(Y\) and a pseudo-metric space \((\mathcal M, \rho)\), if \(F:\mathcal M\to \mathcal K_m(Y)\) satisfies that \(F|_{\mathcal M'}\) has a Lipschitz selection \(f_{\mathcal M'}\) with Lipschitz constant bounded by \(\lambda\) for any \(\mathcal M'\subset \mathcal M\) consisting of at most \(2^{\dim(Y)}\) (resp., \(2^{m+1}\)) points in the case \(\dim(Y)\leq m\) (resp., \(\dim(Y) \ge m+1\)), then \(F\) also has a Lipschitz selection \(f\) with Lipschitz constant bounded by \(\gamma \lambda\), where \(\gamma\) depends only on \(m\). The cases \(m=1,2\) were previously proved by the authors. Also, the case \(Y=\mathbb R^2\) had already been shown by the second named author (see [\textit{P. Shvartsman}, J. Geom. Anal. 12, No. 2, 289--324 (2002; Zbl 1031.52004)]), but the case \(Y=\mathbb R^n\) for metric spaces was still unknown, and it arose in connection with Whitney's extension problem. Observe that, when the result is applied for the trivial distance \(\rho=0\), one obtains Helly's classical theorem, however, as mentioned in the paper, the proof of the general case makes extensive use of Helly's theorem. The result gives sharp ``finiteness constants'' in both the finite- or infinite-dimensional situation.
    0 references
    0 references
    0 references
    set-valued mapping
    0 references
    Lipschitz selection
    0 references
    metric tree
    0 references
    Helly's theorem
    0 references
    Nagata dimension
    0 references
    Whitney partition
    0 references
    Steiner-type point
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references