PARyOpt 1.0.2.1

A flexible Bayesian optimization framework.

This framework performs Asynchronous Bayesian Optimization with support for remote evaluations, resilient to hardware/software failures

Documentation for the software is available on ReadTheDocs.

How do I get set up?

This code is compatible with Python3.5, and requires several modules. The requirements are available in requirements.txt. If you are doing a tar ball installation, do

python3.5 -m pip install -r requirements.txt
python3.5 -m pip install setup.py

If you are using a pip installation, simply do

python3.5 -m pip install paryopt

Citing:

The publication related to the implementation can be found on arxiv. If you use PARyOpt, please cite as:

Pokuri, B. S. S., Lofquist, A., Risko, C. M., & Ganapathysubramanian, B. (2018). PARyOpt: A software for Parallel Asynchronous Remote Bayesian Optimization. arXiv preprint arXiv:1809.04668.

Bibtex entry

@article{pokuri2018paryopt,
title={PARyOpt: A software for Parallel Asynchronous Remote Bayesian Optimization},
author={Pokuri, Balaji Sesha Sarath and Lofquist, Alec and Risko, Chad M and Ganapathysubramanian, Baskar},
journal={arXiv preprint arXiv:1809.04668},
year={2018}
}

Who do I talk to?

Balaji Pokuri

Alec Lofquist

Baskar Ganapathysubramanian