Fundamental Problems at the Interface of Combinatorial Optimization with Integer Programming and Online Optimization

ended

The goal of this proposal is to leverage and significantly extend techniques from the field of Combinatorial Optimization to address some fundamental open algorithmic questions in other, related areas, namely Integer Programming and Online Optimization. More precisely, we focus on the following three thrusts, which share many combinatorial features: - Integer programming with bounded subdeterminants. - Expressive power of mixed-integer linear formulations. - The matroid secretary conjecture, a key online selection problem. Recent significant progress, in which the PI played a central role, combined with new ideas, give hope to obtain breakthrough results in these fields. Many of the questions we consider are long-standing open problems in their respective area, and any progress is thus likely to be a significant contribution to Mathematical Optimization and Theoretical Computer Science. However, equally importantly, if progress can be achieved through the suggested methodologies, then this would create intriguing new links between different fields, which was a key driver in the selection of the above research thrusts.