Evaluation of Optimization Solvers

A comprehensive set of results and problems for evaluating linear programming (LP) and mixed-integer linear programming (MILP) solvers is presented.

Overview

This repository provides results and problems for comparing the performance of various optimization solvers on linear programming (LP) and mixed-integer linear programming (MILP) problems. The evaluation includes metrics such as solution time, quality of solutions, memory usage, and solver reliability across different problem classes. NEOS Server played a key role in achieving these results.

Evaluated Solvers

The repository provides results calculated by the following solvers:

• Open-source solvers:

  • HiGHS
  • SCIP (Solving Constraint Integer Programs)
  • GLPK (GNU Linear Programming Kit)
  • CBC (COIN-OR Branch and Cut)

• Commercial solvers:

  • Gurobi
  • CPLEX
  • XPRESS
  • MOSEK

Problem Sets

The repository includes several benchmark problem sets:

• MIPLIB 2017
• MIPLIB 2010
• Netlib
• Hans Mittelmann's Benchmarks
• Computational Optimization and Applications Software Forum

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Thanks to the developers of all the evaluated solvers.
• Thank you to all the researchers who worked on the benchmarked problems.

Contact

For questions or support, please open an issue or contact [romz@wp.pl]