Kolloquium: Robert Weismantel (ETH Zürich): „Integer optimization in high dimensions"
Integer optimization in high dimensions – Vortragender: Robert Weismantel, Eidgenössische Technische Hochschule Zürich – Einladender: T. Oertel
Abstract: The topic of my talk is integer linear optimization when the dimension is large. The goal is to detect structural results and develop algorithms that are ideally independent on the dimension or depend as little as possible on it. This requires us to assume that the underlying optimization problem has some structure such as bounded subdeterminants or specific sparsity patterns. This will allow us to tackle central questions such as upper bounds on the distance between optimal solutions of a linear programming relaxation and its corresponding integer optimization problem. My talk is based on several papers with different coauthors that will be mentioned in the talk.