Return to Colloquia & Seminar listing
Parametric Cut Generation for Mixed Integer Programs
Student-Run Research| Speaker: | Acadia Larsen, UC Davis |
| Location: | 1147 MSB |
| Start time: | Wed, Oct 14 2026, 2:10PM |
Description
The task of cut selection and generation are important subproblems of branch and cut algorithms for solving mixed integer linear programs. In scip, an open source MILP solver, these subproblems are treated as separate tasks. Gomory's Mixed Integer cuts, an effective family of cuts in practice, can be recognized as intersection cuts generated from a cut generating function. This talk gives a primary report for paramaetricCutGen an open source python package based on sagemath/passagemath that combines which combines tasks of cut selection and generation by solving optimization problems over a space of parameterized cut generating functions. To define optimization problems over a space of parameterized cut generating functions we'll explore combinatorial and geometric properties of the space.
