Mono-X Signal Characterization from Two-component Dark Matter Using a Convolutional Neural Network

Abstract

We assess the scope of a Convolutional Neural Network (CNN) in characterizing potential signals of two-component Dark Matter (DM) arising at the Large Hadron Collider (LHC) from mono-jet and mono-Z probes. We show that such a CNN has the ability of not only inferring the presence of two DM particles but also of extracting their mass and spin, the latter being either 0 or 1/2, following detector level analysis. However, such result represents a conceptual proof-of-concept, as we have not entertained a signal-to-background analysis.