Development has ceased
Unfortunately, due to a new job, I have been unable to find the time to continue developing apsis. If you are still interested in bayesian hyperparameter optimization, you can take a look at, for example, mlrMBO. Should you want to continue this project, feel free to do so, and feel free to contact me for any questions.
A toolkit for hyperparameter optimization for machine learning algorithms.
Our goal is to provide a flexible, simple and scaleable approach - parallel, on clusters and/or on your own machine.
Apsis is open-sourced under the MIT license (see License.txt)
Note that apsis only supports python 2.
We currently are in a beta version. Functionality has been implemented, but has not been used by many people, and there are still advances to tackle.
Our docs are hosted at apsis.readthedocs.org.
We have a mailing list, at apsis-users.
We also have an IRC channel, at irc.freenode.net #apsis.