ingambe/LearningHeuristics

Reinforcement Learning of Dispatching Strategies for Large-Scale Industrial Scheduling

This repository contains the code for the paper "Reinforcement Learning of Dispatching Strategies for Large-Scale Industrial Scheduling"

Install dependencies

pip install -r requirements.txt

If you want to run the linear programming solver, you will need to get and install a Gurobi license. For academics, you can request a free one-year Gurobi license.

Run benchmark

The solver, time-limit and instances used are defined in the run_benchmarks.py file. Once you set up the requested experiments parameters, you can simply run the benchmark script:

python run_benchmarks.py

The results are available in the file results.txt and the file stats.txt.