Dark Matter is extremely prevalent in our universe, making up 85% of the matter in the Universe. One theory about the origin of dark matter is that it exists as a thermal relic of the early universe. The Light Dark Matter Experiment (LDMX) was proposed with the intent to create Dark Matter particles from the interaction of energetic  electrons with nuclei in a fixed target . 

A machine learning algorithm will be very important in helping to discriminate between dark matter signals and known Standard Model (SM) interactions. In this work we utilized two different machine learning algorithms for signal vs background discrimination in LDMX: a boosted decision tree and a deep neural network (ParticleNet). The performance of both approaches is discussed and compared for future use in LDMX.