Related software

Bayesian optimization with Gaussian process

• Limbo (C++11): A lightweight framework for Bayesian and model-based optimisation of black-box functions (C++11). https://github.com/jbmouret/limbo

• pybo (Python): A Python package for modular Bayesian optimization. https://github.com/mwhoffman/pybo

• Spearmint (Python): A library based on [Snoek2012]. It is more oriented to cluster computing. It provides time based criteria. Implemented in Python with interface also to Matlab https://github.com/JasperSnoek/spearmint

• DiceOptim (R): It adds optimization capabilities to DiceKriging. See [Roustant2012] and http://cran.r-project.org/web/packages/DiceOptim/index.html

Bayesian optimization with other models

• SMAC (Java): A library based on [HutHooLey11-smac]. It uses random forests as surrogate model. Intended for hyperparameter optimization. It has Python interface through HPOlib [EggFeuBerSnoHooHutLey13]. http://www.cs.ubc.ca/labs/beta/Projects/SMAC/

• Hyperopt (Python): A library mainly based on [Bergstra2011]. It uses trees of Parzen estimators for the surrogate model https://github.com/jaberg/hyperopt

Related surrogate modeling software

• Perk (Fortran+Shell): A library for surrogate modelling. It was mainly used to illustrate the results in [Santner03]. http://www.stat.osu.edu/~comp_exp/book.html

• SUMO (Matlab+Java): A library for surrogate modelling using Gaussian processes. Its main purpose it is not optimization but it also includes a demo with the Expected Imporvement algorithm. http://www.sumowiki.intec.ugent.be/Main_Page

• GPML (Matlab): The most popular library for Gaussian Process. BayesOpt uses some of its design principles. http://www.gaussianprocess.org/gpml/code/matlab/doc/

Related nonlinear optimization software

• NLOPT (C): One of the best libraries for general purpose nonlinear optimization. It has intefaces for a large number of programming languages. http://ab-initio.mit.edu/wiki/index.php/NLopt