scikinC: a tool for deploying machine learning as binaries
July 14, 2022
Machine Learning plays a major role in Computational Physics providing arbitrarily good approximations of arbitrarily complex functions, for example with Artificial Neural Networks and Boosted Decision Trees. Unfortunately, the integration of Machine Learning models trained with Python frameworks in production code, often developed in C, C++, or FORTRAN, is notoriously a complicated task. We present scikinC, a transpiler for scikit-learn and Keras models into plain C functions, intended to be compiled into shared objects and linked to other applications. An example of application to the parametrization of a detector is discussed.
How to cite
Metadata are provided both in "article" format (very similar to INSPIRE) as this helps creating
very compact bibliographies which can be beneficial to authors and
readers, and in "proceeding" format
which is more detailed and complete.