Answers are grounded in this paper and cite their sources.
Introduction.Helly's intersection theorem [7] asserts that if C is a finite family of convex sets in Rd with HC = 0 then C admits a subfamily K with C=0 and \ gd + 1.The shortest proof, due to Radon [9], is based on the fact that a subset of Rd is affinely independent if and only if it does not contain disjoint sets whose convex hulls intersect.Here a similar approach leads to short proofs of old and new intersection theorems for positive sets.Throughout this note, E denotes a vector space over an ordered field.When E is said to be d-dimensional it should be understood that d is finite.A subset P of E is called positive provided that ax+ßyEP whenever x, y£P and a, ß^O; equivalently, P is a convex cone with apex 0. (When E = Rd, the intersections of positive sets with the unit sphere are precisely the sets which are spherically convex in one of the common meanings of that term.Thus for real vector spaces our theorems could be stated alternatively in terms of spherically convex sets.)The positive hull pos X of a set XEE is the intersection of all positive sets containing X; equivalently, it is the set of all points of the form ¿_iXex with Xx^0for all xandXa; = 0
No takes yet. Share an insight, caveat, or question.
Hansen et al. (1969) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: