Talk:Matroid intersection

intersection of three matroids is NP hard
In this paper: www.math.uni-magdeburg.de/~firla/matching.ps they claim, that there exists matroid intersections of number four, which is polynomial: "The matching problem that is known to be polynomially solvable requires in general at least four matroids." I am not sure if I understand everything correctly, but if they are right we should correct the article.--Flegmon (talk) 12:42, 14 June 2010 (UTC)

I think that what is currently said in the article is correct: even if there may be problems that are polynomially solvable and are representable as the intersection of 3 or more matroids (like for example matchings in arbitrary graphs), this doesn't mean that ALL problems representable by 3 or more matroids can be solve in polynomial time. The intersection problem of three matroids, is in general NP-hard, in particular it contains the asymmetric TSP. 130.149.12.72 (talk) 12:52, 3 September 2010 (UTC) jvers.