pcb442 tsp
- From: Carmel <carmelgafa@xxxxxxxxx>
- Date: Wed, 25 Apr 2007 08:35:55 GMT
Hi all,
I am verifying a tsp solver using GA that I have built and one of the
tests that I have tried is pcb442. What is very strange is that the
optimal solution (that I have found in several websites) does not look
very optimal at all.
In particular the pcb has what it looks like some edge connector on
the left hand side, and instead of visiting each node in one row and
then moving to the next row, the optimal solution is zigzagging from
one row to another.
The GA solution that my program generated (using 2 opt) was better
than the optimal result in all tests carried out.
Can someone tell me if I am missing something? This is the only case
so far where I had such problems.
Thx
Carmel
[ comp.ai is moderated ... your article may take a while to appear. ]
.
- Follow-Ups:
- Re: pcb442 tsp
- From: Carmel
- Re: pcb442 tsp
- From: russell kym horsell
- Re: pcb442 tsp
- Prev by Date: Re: support vector machines
- Next by Date: Re: evolutionary neural networks
- Previous by thread: CFP: IEEE/WIC/ACM WI 2007 (Silicon Valley, USA)
- Next by thread: Re: pcb442 tsp
- Index(es):
Relevant Pages
|
|
