| |
Overview
Parallel/distributed mixed integer programming (PMIP) is a well-known domain
encompassing several critical applications, such as airline crew scheduling.
Many open issues exist in branch-and-cut (B&C) approaches to PMIP, such as
distributed cut-sharing and management in large-scale implementations. In
collaborative research with Operations Research faculty, I have developed a
unique distributed B&C solver called PARINO[ISMP'97,GIT-TR'97]
that facilitates study of many such issues. I am interested in building on this
work and exploring systems issues behind such large-scale cut sharing approaches
(and information sharing in general).
Related Publications
- PARINO, A Parallel Integer Optimizer
Martin Savelsbergh, Kalyan Perumalla, Jeff Linderoth and Umakishore Ramachandran
International
Symposium on Mathematical Programming, August 1997.
[Paper][Ref]
- PARINO: An Extensible Framework for Solving Mixed Integer Programs in
Parallel
Kalyan Perumalla, Martin Savelsbergh and Umakishore Ramachandran
GIT-CC-97-07
College of Computing, Georgia Institute of
Technology, March 1997.
Compressed Postscript
|