Hodge theory for combinatorial geometries (Q1663078)

From MaRDI portal
Revision as of 21:27, 18 April 2024 by Importer (talk | contribs) (‎Changed an Item)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
scientific article
Language Label Description Also known as
English
Hodge theory for combinatorial geometries
scientific article

    Statements

    Hodge theory for combinatorial geometries (English)
    0 references
    0 references
    0 references
    0 references
    21 August 2018
    0 references
    The notion of a matroid, which apriori was introduced as a structure characterizing independence of vector spaces and graphs [\textit{H. Whitney}, Am. J. Math. 57, 509--533 (1935; Zbl 0012.00404)], is of fundamental importance in graph theory. A (finite) matroid \(M\) is given by a pair \((E,\ell)\), where \(E\) is a finite set, and \(\ell\) is a family of subsets of \(E\) (referred to as independent sets) which satisfy certain properties, namely the defining properties of a closure operator and the Mac Lane-Steinitz exchange property. The \textit{chromatic polynomial} \(\chi_G\) associated to a graph \(G\) (which is a function whose value at a positive integer \(q\), \(\chi_G(q)\) is the number of colorings of \(G\) using \(q\) colors) is generalised to the set up of matroids by Rota, and is known as the \textit{characteristic poynomial} associated to a matroid \(M\) [\textit{G.-C. Rota}, Z. Wahrscheinlichkeitstheor. Verw. Geb. 2, 340--368 (1964; Zbl 0121.02406)], defined by \[ \chi_M(\lambda) = \sum_{\ell \subset E} (-1)^{|\ell|} \lambda^{\mathrm{crk}(\ell)} \] where the sum is over all subsets \(\ell \subseteq E\) and \(\mathrm{crk}(\ell)\) is the corank of \(\ell\) in \(M\). A fundamental conjecture that was open in matroid theory, stated in 70's proposes that the absolute values of coefficients of the characteristic polynomial associated to a matroid \(M\) form a log-concave sequence. More precisely, writing \(r + 1\) for the rank of \(M\), which corresponds to the maximum size of an independent set in the matroid, and letting \(w_k(M)\) denote the absolute value of the coefficient of \(\lambda^{r-k+1}\) in the characteristic polynomial of \(M\), the log-concavity would imply: \[ w_{k-1}(M)w_{k+1}(M) \leq w_k(M)^2 ~ \text{ for all } 1 \leq k \leq r. \] A related conjecture which concerns the number of independent subsets of E of given cardinality was formulated by \textit{D. J. A. Welsh} [in: Proceedings of a conference of combinatorial mathematics and its applications, 1969. London: Academic Press. 291--306 (1971; Zbl 0233.05001)] and \textit{J. H. Mason} [``Matroids: unimodal conjectures and Motzkin's theorem'', in: Proceedings of a conference of combinatorial mathematics and its applications, 1972. London: Academic Press. 207--220 (1972)]. The authors prove all these conjectures in full generality by constructing a ``cohomology ring'' associated to \(M\) that satisfies the hard Lefschetz theorem and the Hodge-Riemann relations.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    hard Lefschetz theorem
    0 references
    Hodge-Riemann relation
    0 references
    Bergman fan
    0 references
    matroid
    0 references
    0 references
    0 references
    0 references
    0 references