Automatic event recognition for Higgs boson detection

Several groups of researches have tried, and are still trying, to improve Higgs boson detection. Machine Learning (ML) appears as one of the most promising ways, namely Boosted Decision Trees (BDT), Shallow Neural Networks (SNN), and Deep Neural Networks (DNN). The great advantage of such classifiers is that they can be trained once and then reused several times without needing any significant computational power. Also, they can be free of knowing the physical background, or the meaning of the features. The main aim of this project is to test recently developed ML libraries on data provided by CERN.