Speaker
Description
High-energy particle physics experiments usually involve retrieving useful and interesting event data out of extremely large data sets where in most cases there is less signal and more background. To produce the best possible experimental results it is desirable to optimally discriminate between signal and background. The Toolkit for Multivariate Analysis (TMVA) within ROOT provides many different algorithms for the classification of signal and background events. ATLAS pile-up suppression techniques can impose fake missing transverse energy in the analysis of $E_{T}^{miss}$. In this study, two Multivariate classification techniques were trained to discriminate between signal of real missing transverse energy over background of fake missing transverse energy. The Multivariate techniques used are boosted decision trees (BDTs) and Multilayer perceptron (MLP). The performance of the two classifiers were compared and the BDTs performed better than the MLP classifier.
