In this Masters thesis we have developed a faster way to calculate
supersymmetric cross sections at next-to-leading order (NLO) by using
machine learning techniques. This method teaches the computer software
to imitate the cross section function, facilitating the evaluation of a
large number of parameter points in a short period of time. Training is
carried out based on data generated with SoftSUSY and Prospino 2.1. We
have used the phenomenological MSSM-24 as an example model with the
production of gluino pairs as an example process.