An Inverse Optimization approach to the Amazon Challenge

This repository contains the source code to reproduce the Inverse Optimization (IO) approach to the Amazon Last Mile Routing Research Challenge from the paper Inverse Optimization for Routing Problems.

Usage

To test the code in this repository, you need to follow the steps:

1. Download the Amazon Challenge datasets here.
2. Run process_data.py. This script preprocesses the datasets and splits them per depot (needs to be run only once).
3. Run main.py. This script can be executed in the terminal/command line or IDE. NOTE: before using the main.py file, the variables path_to_input_data and path_to_output_data need to be correctly defined as the path to the processed data and the path to the location where the results will be saved.

The following Python packages are required:

• numpy.
• invopt: this package is used to solve the IO problem.
• gurobipy, ortools or LKH-3: our IO approach to the Amazon Challenge requires a Traveling Salesperson Problem (TSP) solver. Three options are available: a Gurobi-based solver, which solves the TSP to optimality, but is possibly slow for large TSPs; or LKH-3 or Google OR-Tools, which solves the TSP approximately, but are possibly faster for large TSPs. NOTE: to use the LKH-3 solver, the variable solver_path in the utils.py file needs to be correctly defined.

Citing

If you use this repository, please cite the accompanying paper:

@article{zattoniscroccaro2024inverse,
  title={Inverse Optimization for Routing Problems},
  author={Zattoni Scroccaro, Pedro and van Beek, Piet and Mohajerin Esfahani, Peyman and Atasoy, Bilge},
  journal={Transportation Science},
  year={2024}
}