|
Over the years, powerful software tools been developed to tackle integer
programming (IP) problems. IP has applications in many areas: For example, crew
scheduling, transportation optimization, supply-chain management etc.
With the availability of mature and robust IP solvers, one is tempted (or even
expected) to look even further for problems for which one can use such solvers to
obtain solutions.
In this talk, two contrasting problems will be considered. One is from an area
that has little to do with IP at first glance. The other is a classic problem in
combinatorial optimization. The IP formulations for both problems pose little
difficulty for modern-day solvers yet whether or not one likes what the solvers
return can depend on how one feels on a given day. Suggestions on how one could
address such an uncertainty will be given.
|