GPFlowOpt

Library for Bayesian optimization and active learning (only for Tensorflow 1.x)

Description:

A Tensorflow 1 library (See Trieste for the version for the latest Tensorflow library) with the latest Bayesian optimization strategies. Bayesian optimization can be used for efficient hyperparameter tuning (AutoML), smart labeling of expensive data, data-efficient engineering design, and much more.

More Information:

Originates from the AI-Driven Data Science research challenge (FAIR 2019–2023).
Responsible Research Lead: Tom Dhaene
Authors: Ivo Couckuyt
Programming Language: Python
Link to the website
Link to the repository
Categories: Optimization, Active learning

Publications:

N. Knudde, J. van der Herten, T. Dhaene, I. Couckuyt, "GPflowOpt: A Bayesian Optimization Library using TensorFlow", arXiv:1711.03845v1 [stat.ML] , 2017