Re: Can this be solved in polynomial time?
- From: daw@xxxxxxxxxxxxxxxxxxxxxxxx (David Wagner)
- Date: Sat, 14 Apr 2007 01:25:22 +0000 (UTC)
You have a set S of size N.
You are given M subsets of S.
Find K of the given subsets, so that their union is of maximal size.
Can this problem be solved efficiently? If so, how?
This is not homework.
Ok. Thanks.
I do not think it can be solved efficiently. Proof by reduction to the
set covering decision problem. WLOG we can assume that the union of
the M subsets is all of S. If you could solve your optimization problem
quickly, you could also answer the following question quickly:
Does there exist K of the given subsets, whose union is all of S?
(Just run your algorithm for your optimization problem, and see whether
the K subsets it produces cover all of S.) But the latter question
is exactly the set covering decision problem, which is known to be
NP-complete.
.
- Follow-Ups:
- Re: Can this be solved in polynomial time?
- From: Leo2157
- Re: Can this be solved in polynomial time?
- References:
- Can this be solved in polynomial time?
- From: Leo2157
- Re: Can this be solved in polynomial time?
- From: David Wagner
- Re: Can this be solved in polynomial time?
- From: Leo2157
- Can this be solved in polynomial time?
- Prev by Date: Re: Linear programming?
- Next by Date: Hardness of this problem
- Previous by thread: Re: Can this be solved in polynomial time?
- Next by thread: Re: Can this be solved in polynomial time?
- Index(es):