GUI is based on TSP visual: https://github.com/bcyran/tsp-visual

In this project we tried to come up with a solution for a complex minimax problem, considering a TSP variant which tries to maximize revenues and minimize independent expenses.

Implemented algorithms:

• Genetic algorithm
• Greedy search baseline
• Brute force - optimal (w.r.t score)

instalation and running:

pip install -r requirements.txt
python -m noa_kirel

Using TSP Visual is pretty straightforward, using drop-down menus will load a pre-generated dataset from noa_kirel/datasets directory; same goes for solvers. Simply choose your desired solver and dataset, adjust the hyper parameters and hit "Solve".

• Implemented solvers could be found under noa_kirel/solvers
• Evaluate: run noa_kirel/evaluation.py and then noa_kirel/analyze_results.py; Note that this may take some time!