Concepts of Dimension for Convex Geometries
Kolja Knauer,William T. Trotter
DOI: https://doi.org/10.1137/23m1559853
IF: 0.868
2024-05-24
SIAM Journal on Discrete Mathematics
Abstract:SIAM Journal on Discrete Mathematics, Volume 38, Issue 2, Page 1566-1585, June 2024. Let [math] be a finite set. A family [math] of subsets of [math] is called a convex geometry with ground set [math] if (1) [math]; (2) [math] whenever [math]; and (3) if [math] and [math], there is an element [math] such that [math]. As a nonempty family of sets, a convex geometry has a well defined [math]-dimension. In the literature, a second parameter, called the convex dimension, has been defined expressly for these structures. Partially ordered by inclusion, a convex geometry is also a poset, and four additional dimension parameters have been defined for this larger class, called the Dushnik–Miller dimension, Boolean dimension, local dimension, and fractional dimension, respectively. For each pair of these six dimension parameters, we investigate whether there is an infinite class of convex geometries on which one parameter is bounded and the other is not.
mathematics, applied