Seminar on Scheduling and Decomposition
a.Angelegt durch Lehrenden
Akt: 25.10.2021 11:33
In this seminar, we will study variants of decomposition algorithms applied to scheduling and inventory location problems.
Scheduling problems are considered particularly hard for MIP solvers.
For this reasons, dynamic programming or constraints programming techniques are often applied instead.
However, applications in industry frequently complicate these problems further as they introduce additional constraints and objectives that render these non-MIP based approaches infeasible.
This issue can be addressed by exact or heuristic decomposition algorithms.
Here, we aim to decompose the problem into master and subproblems each exhibiting well-studied structure.
These problems are then solved repeatedly with appropriate algorithms.
In each iteration, the solution of the subproblems is used to guide search in the master problem and vice versa.
This seminar is held as a block seminar at the end of the semester.
The full list of papers and abstracts is available in the whiteboard.
To participate, you should have a solid background in discrete math and linear programming equivalent to the lectures "Discrete Math I" and "Optimization I" at FU, or "ADM I" at TU.
02.11.2021 14:00 - 16:00