Re: Is this proof right?
From: Alexander T. Hamming (ath77_at_aol.com)
Date: 10/28/04
- Next message: Russell Wallace: "Re: Largest provable BB(N)?"
- Previous message: Robert Low: "Re: Largest provable BB(N)?"
- In reply to: SD: "Re: Is this proof right?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: 28 Oct 2004 10:11:10 -0700
fake.addr@gmail.com (SD) wrote in message news:<10cddc5b.0410241448.53b646c4@posting.google.com>...
> Casey Hawthorne <caseyhHAM@SCRAMistar.ca> wrote in message news:<junmn0tcci1sqc1685mlom5rftdpd1lrbk@4ax.com>...
> > I believe that for the TSP to be solved in polynomial time by linear
> > programming the cities must be on lattice points.
>
> From what I read in the paper it required $x_{i s j} \in \{ 0, 1 \} $
> which, from what I recall, means we are dealing with a integer
> programming problem. Therefore it is not an LP formulation for TSP.
No! he made an IP formulation of TSP. Then he relaxed the constraints
so 0<=x<=1, making it an LP problem. He claimed that the maximum
solution to the LP problem is always integer with $x_{i s j} \in \{ 0,
1 \} $. I want to know if this is a correct claim. If so and if the IP
formulation is correct, then NP=P.
- Next message: Russell Wallace: "Re: Largest provable BB(N)?"
- Previous message: Robert Low: "Re: Largest provable BB(N)?"
- In reply to: SD: "Re: Is this proof right?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|