[Dip] Benders decomposition in DIP
Jonas Christoffer Villumsen
jcvi at man.dtu.dk
Fri May 6 02:25:45 EDT 2011
Thanks Jiadong. I will consider using AMPL instead.
I was just wondering - since DIP supports cutting plane algorithms - as far as I understand - it might also be possible to implement an algorithm using Benders decomposition in a similar way(?).
Jonas
From: Jiadong Wang [mailto:jiadwang at gmail.com]<mailto:[mailto:jiadwang at gmail.com]>
Sent: 5. maj 2011 13:32
To: Jonas Christoffer Villumsen
Cc: dip at list.coin-or.org<mailto:dip at list.coin-or.org>
Subject: Re: [Dip] Benders decomposition in DIP
Hey Jonas
As far as I know, DIP focuses on Dantzig-Wolfe decomposition. The reason might be that there are some critical applications which can be tailored to column generation methods in OR field.CG is more difficult to implement in terms of transforming between different spaces for the convexification approach.
Indeed, there are some applications which have many constraints can be more easily solved by Benders decomposition.
Also convergence issues might be different for these two decompositions.
there is some example http://www.ampl.com/NEW/LOOP2/index.html showing how to build benders decomposition in AMPL.
Matt might add more...
Jiadong
On Thu, May 5, 2011 at 6:44 AM, Jonas Christoffer Villumsen <jcvi at man.dtu.dk<mailto:jcvi at man.dtu.dk>> wrote:
Hi all,
Is it possible to implement a Benders decomposition approach in DIP? And if so, how would one go about this?
If anyone has any experience or ideas for this, I would greatly appreciate your comments. Thanks!
Best regards,
Jonas
_______________________________________________
Dip mailing list
Dip at list.coin-or.org<mailto:Dip at list.coin-or.org>
http://list.coin-or.org/mailman/listinfo/dip
--
Jiadong Wang
Harold S. Mohler Laboratory
Lehigh University
200 West Packer Avenue
Bethlehem, PA 18015-1582
Phone No: 484-707-8895
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.coin-or.org/pipermail/dip/attachments/20110506/2fef4711/attachment.html
More information about the Dip
mailing list