CakeOpt is a Bayesian optimisation package written in Python. Features include support for noisy function evaluations and mixed parameter spaces.
CakeOpt requires Numpy, Scipy, Matplotlib, Scikit-Learn, and GPy. From a terminal with a working Python installation:
pip install numpy scipy matplotlib sklearn gpy
git clone [email protected]:m-t-m/CakeOpt.git
CakeOpt implements algorithms described in the dissertation below:
Bayesian Optimisation over Mixed Parameter Spaces
Marius Tudor Morar
Doctoral dissertation, University of Manchester, 2021