SUSY-AI is hosted by Hepforge, IPPP Durham

SUSY-AI: The Supersymmetric Artificial Intelligence Package

SUSY-AI is a python program which determines if a benchmark point is excluded or not at the 95 % C.L. by comparing to LHC data in the MSSM context. It uses machine learning classification techniques to provide a fast answer. Currently there are two SUSY-AI versions, each working onthe phenomenological MSSM: one for checking against 8 TeV exclusion limits (see here) and another one for checking against 13 TeV exclusion limits (see here ). SUSY-AI can be used in investigating the exclusion of a model point in the full pMSSM, but also in its submodels, such as natural SUSY, the cMSSM and many other SUSY scenarios.


If you use SUSY-AI for your publication, please refer to the following paper arXiv:1605.02797.


The code has some external dependences in form of pickle files that can be downloaded from downloads as well as the susyai package. Details about how to setup the package and its usage can be found in the manual placed in the documentation.

Important All classifiers on this website have been created with scikit-learn version 0.16. Due to a change in internal structure of the objects in scikit-learn version 0.18.1, the classifiers won't load in versions >= 0.18.1.


The code is maintained by Sascha Caron, Jong Soo Kim, Krzysztof Rolbiecki, Roberto Ruiz de Austri and Bob Stienen.