Using ML techniques to discriminate the tHq(b ¯b) decay channel signal from background

Machine Learning techniques are of very importance when analyzing large amounts of data as the ones acquired by the ATLAS detector. This project focus on the employment of Graph Convolutional Networks (GCN) together with the Deep Graph Library (DGL) to discriminate the signal of the production of a Higgs boson and a single-top quark in the tHq(b ¯b) channel . Python scripts were written to create the graphs and to train and test the Neural Networks. The main goal of using DGL to create the GCN was achieved, however several parameters still have to be altered in order to get a higher test accuracy.