Primary Vertex Selection in Higgs to 2 Photon Events

Determining the position of the primary vertex that produces a Higgs boson is critical to maximize our sensitivity to a Higgs signal. In H → γγ decays the vertex reconstruction allows us to calculate the photons direction accurately, and hence the di-photon invariant mass resolution. The challenge is to select the correct one among the multiple possible positions. In our work, we try to use some machine learning methods to improve the accuracy and efficiency of selection. Furthermore, this project will investigate new options that will be optimal now (Run 2/Run 3) and in the future (HL-LHC) using additional information from the reconstructed event. So far, when using the random forest plus AdaBoost, we get the best accuracy rate of 96.5%.