Bandit-based-HPO

Towards Bandit-based Optimization for Automated Machine Learning

This repository contains the implementation of the paper "Towards Bandit-based Optimization for Automated Machine Learning, accepted at ICLR 2024 Workshop on Practical Machine Learning for Low Resource Settings (PML4LRS) by Amir Rezaei Balef, Claire Vernade and, Katharina Eggensperger"

Dependency

Using a Conda environment is recommended.

You may need to install and set up the TabRepo and YAHPO gym packages.

TabRepo: https://github.com/autogluon/tabrepo

YAHPO gym: https://github.com/slds-lmu/yahpo_gym

To install the repository, ensure you are using Python 3.9-3.11. Other Python versions may not be supported. Then, run the following commands:

git clone https://github.com/amirbalef/Bandit-based-HPO
pip install -r requirements.txt

Only Linux support has been tested.

Running expiriments

To run experiments, execute the following command:

python main.py 

Feel free to adapt and extend this codebase as needed for your own experiments and research.