Duality theory for finite and infinite matroids with coefficients (Q1109782): Difference between revisions

A standard example for a matroid is given by a presentation via some vector subspace V of \(K^ E\) (K a field, E a finite set) in the sense that the circuits of the matroid presented in this fashion are the minimal nonvanishing supports of the vectors in V. This approach to matroids is put into a very general algebraic framework designed to provide a unifying treatment of representable and nonrepresentable matroids as well as oriented matroids. The role of the field K is taken over by a so-called fuzzy ring, which one may think of as a ring factored by a subgroup of its group of units. The subspace V then becomes a set of relations so that a matroid defining condition holds. In this way, matroids can also meaningfully be defined with respect to (possibly infinite) matroid support systems which are required to be closed under a Galois connection based on finite intersection of sets. The dual matroid may now be defined with respect to the Galois dual of the support system and it is shown that the dual of the dual yields the original matroid. Furthermore, for the lattice of flats of such a matroid, properties enjoyed by geometric lattices are verified and the rank function is investigated.