|Description:||The particle accelerators at CERN produce a high number of so-called events, describing the collision products and its properties. The project is a part of the effort to analyse the properties of the Higgs boson. Prior to their deployment, novel machine learning based methods for classifying corresponding events have to be trained and evaluated on data obtained from simulating the ATLAS detector. The goal of the thesis research is to develop a generative model (e.g. variational autoencoder) for enhancing the available dataset of Higgs boson related events. The project offers the possibility to be part of an international collaboration with a visit to CERN. Details of the project will be clarified by consultation with the supervisor.