Some simple python tools for solving the Minimum-Feature-Set problem.
Exact solvers using:
- CPLEX (best performance)
- localsolver (horrible performance)
- numberjack (quite poor)
- Googles ortools (good performance but slower than with CPLEX)
Heuristic solvers:
- Greedy (Fast but often suboptimal)
- Meta-RaPS (Slower but scales much better than the exact solvers and with better results than the greedy heuristic)
The Meta-RaPS heuristic reduces the problem to the equivalent Set Cover Problem and solves with this method.