Two curses affecting Mixed-integer programming computation

In today‚Äôs last post dedicated to the #MIP2013 workshop, I examine two sources of complexity affecting research in mixed-integer (linear and nonlinear) programming. Curse of Dimensionality Although the term was coined by Richard E. Bellman in context of dynamic programming, this applies to most aspects of integer programming. Even a LP-based polytope has an exponential […]

Closing the (MIP) Gap – Part II

Day #2 of the MIP 2013 Workshop reminded me that despite the numerous research efforts invested in mixed-integer programming in the last 60 years or so, much remains to be done. A lot of MIP models are still very difficult to solve. In MIP extensions such as nonlinear, multi-objective or multi-level programming, most large-sized problems […]