Abstract
The Higgs boson couplings to bottom and top quarks have been measured and agree well with the Standard Model predictions. Decays to lighter quarks and gluons, however, remain elusive. Observing these decays is essential to complete the picture of the Higgs boson interactions. In this work, we present the perspectives for the 14 TeV LHC to observe the Higgs boson decay to gluon jets assembling convolutional neural networks, trained to recognize abstract jet images constructed embodying particle flow information, and boosted decision trees with kinetic information from Higgs-strahlung events. We show that this approach might be able to observe Higgs to gluon decays with a significance of around 2.4σ improving significantly previous prospects based on cut-and-count analysis. An upper bound of BR(H → gg)≤1.74 × BR SM (H → gg) at 95% confidence level after 3000 fb−1 of data is obtained using these machine learning techniques.
Highlights
The Standard Model (SM) Higgs boson established the spontaneous electroweak symmetry breaking (EWSB) mechanism as the responsible to give the particles their masses at the same time that it preserves the gauge symmetry of the SM interactions [1–5]
The LHC is confirming that the Higgs boson coupling scales with the mass of the SM particles
This makes it easier to probe the Higgs couplings to bottom and top quarks, the heavy gauge bosons and the tau lepton but it turns the observation of interactions to the light quarks, especially up, down and strange quarks, and the lighter leptons, much more challenging
Summary
The Standard Model (SM) Higgs boson established the spontaneous electroweak symmetry breaking (EWSB) mechanism as the responsible to give the particles their masses at the same time that it preserves the gauge symmetry of the SM interactions [1–. Different approaches to increase the amount of information contained in a jet-image have been proposed [26, 31, 42] These methods employ a hybrid use of high-level features (i.e. kinematic observables) together with the information recovered from each detector section (electromagnetic calorimeter (ECAL), hadronic calorimeter (HCAL), Muon chambers, tracking system) encoded as image channels, greatly improving the amount of information, which subsequently increase the discriminant power of the algorithms. This improvement comes at the price of drastically increase of the model complexity which leads to overfitting and/or a slow training phase for the NN. We ensure that the transformations are not applied two times consecutively in the same image, so that if an image has first an horizontal flip, the transformation will not be flipped horizontally again
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
