Skip to content

Implementation of the Expected Improvement and the Gaussian Process Upper Confidence Bound algorithm in MATLAB, as part of my Bachelor thesis @ ETHZ in 2014.

Notifications You must be signed in to change notification settings

cglanzer/bayesian-global-optimization

Repository files navigation

Bayesian Global Optimization Using Gaussian Processes

Bachelor Thesis, 2014, ETHZ

This repository contains implementations of the Expected Improvement and the Gaussian Process Upper Confidence Bound algorithm in MATLAB, which are part of my Bachelor thesis.

For a documentation of the code, please read Chapter 5 of my thesis here. A complete version of my thesis can be found on my website.

To run the simulations, you need a recent version of MATLAB, including both the Optimization Toolbox and the Statistics Toolbox. The file simulation.m is a template showing how to run the algorithms on specific functions. The template is self-explaining and well commented. For details, please take a look at EI.m and GPUCB.m. Each function is commented and has a header which explains the parameters.

About

Implementation of the Expected Improvement and the Gaussian Process Upper Confidence Bound algorithm in MATLAB, as part of my Bachelor thesis @ ETHZ in 2014.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages