A Flow-Based Solver for Large-Scale Combinatorial Optimization
Alexander Souza ⋅ Lucien Kern ⋅ Tiago Matos ⋅ Kai Schärer ⋅ Tristan Koning
Abstract
We report on our design, implementation, and computational results of a new flow-based solver for combinatorial optimization. The technology is capable of solving very large instances having up to billions of graph-edges within days and accuracy of $1-2\%$, consistently. The results have been validated on real-world instances occurring in the Swiss railway industry.
Chat is not available.
Successful Page Load