On September 10, 2026, our colleague Gioni Mexi (AISST/IOL) successfully defended his doctoral dissertation titled "The Two Faces of Mixed-Integer Programming: Primal and Dual Progress" at the Technische Universität Berlin.
In his research, Gioni Mexi works on the two fronts of mixed-integer programming, finding good solutions quickly and learning from infeasible subproblems. On the primal side, he develops two heuristics that find high-quality solutions efficiently. On the dual side, he advances conflict analysis, which turns encountered infeasibilities into constraints that guide the remaining search, and introduces a new branching rule.
His methods have been recognized at international conferences and solver competitions, and are available as open-source tools and integrated into the solver SCIP. The research was conducted and funded as part of the SynLab at Research Campus MODAL. We warmly congratulate him on his successful defense!
