LamNgo1 / cma-meta-algorithm

[TMLR' 24] High-dimensional Bayesian Optimization via Covariance Matrix Adaptation Strategy

Overview

Code for the CMA-Meta-Algorithm for Bayesian Optimization from High-dimensional Bayesian Optimization via Covariance Matrix Adaptation Strategy in the Transactions on Machine Learning Research (TMLR). The OpenReview page can be found in this link.

If you find our paper or this repo to be useful for your research, please consider citing:

@article{ngo2024cmabo,
  title={High-dimensional Bayesian Optimization via Covariance Matrix Adaptation Strategy},
  author={Ngo, Lam and Ha, Huong and Chan, Jeffrey and Nguyen, Vu and Zhang, Hongyu},
  journal={Transactions on Machine Learning Research},
  year={2024}
}

Requirement and dependencies

• Install Anaconda/Miniconda.
• Install all dependencies listed in environment.yml file, or run the following to install via conda:

conda env create -f environment.yml

• If you want to run Lasso Benchmark, install the LassoBench library by:

cd test_functions/
pip install -e .

Getting started

The main file is test-main.py. To run the code for any problem, use the following syntax:

python test-main.py --solver <optimizer_name> -f <function_name> -d <input_dim> -n <number_of_iterations> --seed <seeding>

Available options for arguments:

• optimizer_name: bo, turbo, baxus
• function_name: alpine, levy, ellipsoid, rastrigin, shifted-alpine, shifted-levy, ellipsoid, rastrigin, branin500, schaffer100, lasso-dna, rover100, half-cheetah
• input_dim: an integer (only applicable for alpine, levy, ellipsoid, rastrigin, shifted-alpine, shifted-levy)
• number_of_iterations: an integer to describe to iteration budget.
• seeding: (optional) the random seed for random behavior of random, numpy, and torch.

Results from paper

We provide the settings for reproducing results presented in the paper experiments-from-paper.sh.

Acknowledgements

This implementation uses materials from the following public repositories to implement the CMA local regions and the incorporated BO optimizers. We thank the respective repository maintainers.

1. CMA-ES: Hansen, N., & Ostermeier, A. (2001). Completely derandomized self-adaptation in evolution strategies. Evolutionary computation, 9(2), 159-195. Code repo: https://github.com/CMA-ES/pycma
2. TuRBO: Eriksson, D., Pearce, M., Gardner, J. R., Turner, R., & Poloczek, M. (2019). Scalable global optimization via local Bayesian optimization. Advances in Neural Information Processing Systems, 32 (NeurIPS). Code repo: https://github.com/uber-research/TuRBO/
3. BAxUS: Papenmeier, L., Nardi, L., & Poloczek, M. (2022). Increasing the scope as you learn: Adaptive Bayesian optimization in nested subspaces. Advances in Neural Information Processing Systems, 35, (NeurIPS). Code repo: https://github.com/LeoIV/BAxUS