Symmetries arise naturally in integer programs. Often, there is no reason to distinguish certain objects among others. For instance, the actual number that a certain bus line receives is irrelevant for the quality of the schedule. In the resulting models, a possibly large symmetry group operates on the set of feasible solutions, for which the respective objective function is constant on every orbit. Algorithms that do not take such symmetries into account very often perform poorly. The goal of this project is to develop structural insights and algorithmic methods as well as software tools to enhance today's integer programming solving technology. Our project vision is to benefit rather than suffer from symmetries in integer programming models.

Publications

2008
Constraint Integer Programming: Techniques and Applications ZIB-Report 08-43 Tobias Achterberg, Timo Berthold, Stefan Heinz, Thorsten Koch, Kati Wolter PDF
BibTeX
URN
Symmetries in Integer Programming
Detecting Orbitopal Symmetries ZIB-Report 08-33 (Appeared in: Operations Research Proceedings 2008, Bernhard Fleischmann ... (eds.), 2009, pp. 433-438) Timo Berthold, Marc Pfetsch PDF
BibTeX
URN
Symmetries in Integer Programming
2006
Orbitopal Fixing ZIB-Report 06-48 (Appeared in: Proc. of the 12th Integer Programming and Combinatorial Optimization Conference (IPCO) M. Fischetti and D. Williamson (eds.), LNCS 4513, Springer-Verlag, 74-88) Volker Kaibel, Matthias Peinhardt, Marc Pfetsch PDF
PDF
BibTeX
URN
Symmetries in Integer Programming
Packing and Partitioning Orbitopes ZIB-Report 06-17 (Appeared in: Mathematical Programming 114 (2008) 1-36) Volker Kaibel, Marc Pfetsch PDF
BibTeX
URN
Symmetries in Integer Programming