Phenomenology of U(1)<em>_{L_μ-L_τ}</em> Gauge Extension of the Standard Model at LHC

Abstract

Extending the Standard Model (SM) by a U(1)_{L_mu-L_tau} group gives potentially significant new contributions to g_mu-2, allows the construction of realistic neutrino mass matrices, incorporates violation of lepton universality violation, and offers an anomaly free mediator for a Dark Matter (DM) sector. In this thesis, we focus on constraints on this model from the Large Hadron Collider (LHC) data based on final states with two, three, or four leptons and missing energy. On the one hand, We recast a large number of LHC analyses involving relevant final states using the CheckMATE framework; and set the upper LHC limit. However, most of the existing LHC analyses are not specially designed for this model, which leads to a possibility of optimizing the cuts. Therefore, on the other hand, we apply several Machine Learning (ML) algorithms to distinguish the model from the SM using simulated LHC data; and derive the bound accordingly. We also try to uncover the "black box" of ML by discussing the feature importance, which is extracted from Gradient Boosting Decision Tree (GBDT). The same scheme could be generalized to other models.