Freedom251 › Statistical Analysis Software
BayesO
What BayesO is
BayesO is a free, MIT-licensed Python package for Bayesian optimization on Linux. It is intended for global optimization when a function behaves as a black box or takes significant resources to evaluate, and it can support both straightforward projects and application-specific changes. The package provides Gaussian process regression and Student-t process regression as surrogate models. Its acquisition choices include probability of improvement, expected improvement, Gaussian process upper confidence bound, log expected improvement, augmented expected improvement, pure exploitation, and pure exploration. Documentation demonstrates tuning max_depth and n_estimators for an XGBoost classifier and optimizing a function with WandB. BayesO supports scripting and is tested with Python 3.8 through 3.12. Installation is available through PyPI with pip install bayeso or from source, including editable development mode. Optional, development, and example dependencies are available through extras.
Who it suits
BayesO suits Python users working on Bayesian optimization, including black-box objectives, expensive evaluations, and application-specific modifications. It is also relevant to users tuning machine-learning hyperparameters or integrating optimization with WandB.
What it does well
- Free MIT-licensed package
- Gaussian process and Student-t surrogates
- Several acquisition-function choices
- Python 3.8 through 3.12 tested
- PyPI and source installation
What to know first
- Linux platform listed
- No free trial
- Optional dependencies require extras
Plans
| Plan | Billed and limits | Price |
|---|---|---|
| BayesO | MIT-licensed Python package | Free |
As bayeso.readthedocs.io lists them, read 12 Oct 2026. The maker's own page has the final word.
Verdict
BayesO provides a focused Bayesian optimization toolkit with multiple surrogate and acquisition options. Its Linux and Python-package format fit scripting and development workflows, while the examples show machine-learning and WandB applications.
Facts from BayesO's own pages
- Acquisition functions
- Included acquisition functions include probability of improvement, expected improvement, and Gaussian process upper confidence bound. · bayeso.readthedocs.io, 12 Oct 2026
- Bayesian analysis
- Yes · bayeso.readthedocs.io, 23 Sept 2026
- Free plan
- Yes · bayeso.readthedocs.io, 23 Sept 2026
- Hyperparameter optimization
- An example uses BayesO to optimize max_depth and n_estimators for an XGBoost classifier. · bayeso.readthedocs.io, 12 Oct 2026
- Installation
- The package can be installed from PyPI with pip install bayeso or from source, including in editable development mode. · bayeso.readthedocs.io, 12 Oct 2026
- Integrations
- The documentation includes a script and example for optimizing a function with WandB. · bayeso.readthedocs.io, 12 Oct 2026
- Intended users
- The documentation describes use for application-specific modifications and for running Bayesian optimization in various applications. · bayeso.readthedocs.io, 12 Oct 2026
- License
- BayesO is licensed under the MIT License. · bayeso.readthedocs.io, 12 Oct 2026
- Optimization use
- Bayesian optimization is presented as a strategy for global optimization of black-box and expensive-to-evaluate functions. · bayeso.readthedocs.io, 12 Oct 2026
- Optional packages
- Optional, development, and example dependencies can be installed with the optional, dev, and examples extras. · bayeso.readthedocs.io, 12 Oct 2026
- Other acquisition options
- The API also lists log expected improvement, augmented expected improvement, pure exploitation, and pure exploration. · bayeso.readthedocs.io, 12 Oct 2026
- Purpose
- BayesO is a Python package for Bayesian optimization, designed for advanced or straightforward use across applications. · bayeso.readthedocs.io, 12 Oct 2026
- Scripting support
- Yes · bayeso.readthedocs.io, 23 Sept 2026
- Security and compliance
- The opened documentation pages do not state security or compliance certifications. · bayeso.readthedocs.io, 12 Oct 2026
- Supported Python
- The project says it tests Python versions 3.8 through 3.12. · bayeso.readthedocs.io, 12 Oct 2026
- Surrogate models
- The package includes Gaussian process regression and Student-t process regression surrogate functions. · bayeso.readthedocs.io, 12 Oct 2026
Where BayesO stands on Freedom251
- Statistical Analysis Software#71 of 106