Santa's Elf Scheduling Genetic Algorithm Challenge

Overview

Welcome to the GA Elf Scheduling Challenge! The genetic algorithm tutorial and challenge details appear in genetic_algorithm_challenge.ipynb

Getting Started

1. Run pip install -r requirements.txt to install required packages.

2. The code defining the problem is in src/problem.py and the algorithm testing code is in opt_experiment.py

3. You are only required to edit the files src/genetic_algorithm_EDIT.py and src/opt_params_EDIT.py for this challenge in the sections marked # TODO: ADD CODE HERE.

4. You may optionally also overwrite the generate_children() function inherited from src.base.BaseGA.

5. You may also write your own script to conduct hyperparameter optimisation on the GA's parameters.

Solution

Solution on solution branch (don't peek!), including ga-challenge-2022-debrief.pdf showing the performance of different GA variants over time.

The following files should be run when testing:

• opt_experiment_base.py - Run an optimisation experiment for the base GA
• opt_experiment_improved.py - Run an optimisation experiment for the improved GA
• optuna_experiment.py - Conduct HPO on the parameters of the improved GA
• utils.py - Generate a graph comparing the average performance of the different algorithms over time

Note: Benchmarking was conducted on a MacBook Pro, macOS Monterey 12.5.1 | 2 GHz Quad-Core Intel Core i5 | 16 GB 3733 MHz LPDDR4X