Speaker
Description
The Dark Matter question is among the most fundamental unsolved problems of modern day physics. One possible candidate for Dark Matter are Dark Mesons, a type of Stealth Dark Matter. They are predicted by extending the Standard Model with a strongly coupled dark sector. Similar to quarks they appear in confined states like baryons and mesons. Their coupling to the Higgs field allows interaction with Standard Model particles that could lead to their production at the Large Hadron Collider at CERN and their indirect detection in the ATLAS experiment. For distinguishing between the signatures of processes involving Dark Mesons and already known Standard Model processes a set of variables describing the composition and kinematics of the final states are used. The information that is extracted from these variables can be exploited even more efficiently if used in combination with machine learning. Boosted Decision Trees can be trained to make elaborate tree-shaped systems of cuts that will concatenate the discriminating power of many variables into one single final variable. The extent to which this method can improve the ATLAS search for Dark Mesons is studied in this masters project.
