Is it specified anywhere what the backend is for the Bayesian optimization related functionality in Mathematica v11? I'm specifically referring to the Bayesian*
functions like BayesianMinimization
. Are they based on an open-source package like Spearmint, or on a brand new implementation? For example the neural network functionality is based on MXNet, but I couldn't find any information in the docs about the backend that's being used for Bayesian optimization.
$\begingroup$
$\endgroup$
1
-
1$\begingroup$ it seems to use the same implementation of Gaussian Processes as is used by the Predict/Classify framework. this suggests it's probably a brand new implementation. $\endgroup$– Michael CurryCommented Jan 17, 2018 at 2:35
Add a comment
|