Explore chapters and articles related to this topic
Biobjective optimization problems on matroids with binary costs
Published in Optimization, 2023
Jochen Gorski, Kathrin Klamroth, Julia Sudhoff
Recall from Section 3 that the objective values of the binary objective b can only take values between 0 and m, i.e. , where m is the rank of the underlying matroid and B is an arbitrary basis. As a consequence, we have that . For an instance of the graphic matroid on a connected graph with n vertices and m edges, this implies that for all spanning trees T of G. Note that due to the common notation that and for graphic matroids, the rank of a graphic matroid is thus n−1. The CE approach determines all efficient spanning trees by maintaining a list with n entries, one for each potential value of that stores the currently best cost value together with all corresponding trees that were enumerated so far.