home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.theory
- Path: sparky!uunet!caen!zaphod.mps.ohio-state.edu!rpi!chaudhus
- From: chaudhus@rolls.cs.rpi.edu (Samit Chaudhuri)
- Subject: Re: The P=NP question (was Re: NP Completeness Question
- Message-ID: <0ay1n4k@rpi.edu>
- Nntp-Posting-Host: rolls.cs.rpi.edu
- Organization: Rensselaer Polytechnic Institute, Troy, NY
- References: <Bxpv45.2I1@watserv1.uwaterloo.ca>
- Date: Wed, 18 Nov 1992 22:38:30 GMT
- Lines: 19
-
- The Research Report CIS92-D4 by Gismondi, S.J. and Swart E. R. that gives a
- polynomial time algorithm to the graph isomorphism problem doesn't make a
- point very clear:
-
- In Theorem 1 it proves that every extreme point of the formulation F3
- corresponds to,
- a) An extreme point of C0, the convex hull of all the doubly
- stochastic Q matrices (Note that they do not necessarily represent an
- isomorphism between the graphs under consideration. The additional
- constraints Ta = b are required to ensure that.), or
- b) A convex combination of an extreme point of C0 and a point in
- C1 (the feasible region of formulation F1)
-
- In Corollary 1 it says that the same is also valid for the formulation F4
- (F4 satisfies all the constraints of F3 together with the additional
- constraint Ta = b). Would the new constraint necessarily preserve the
- properties of F3 ?
-
- It didn't seem obvious to me, could anyone explain ?
-