A Realistic and Efficient Boosted Decision Tree in LDMX

The Light Dark Matter eXperiment (LDMX) is a fixed target accelerator experiment searching for dark matter in the sub-GeV mass range. From the experiment's Electromagnetic Calorimeter (ECal) hits, specific variables are calculated, and fed to a Boosted Decision Tree (BDT). The BDT is used in LDMX to distinguish between signal and background events, and is crucial to the experiment's long term goal. Previously, tracks from Minimum Ionizing Particles (MIPs) have been used in the BDT, but are no longer central to background rejection. Calculating these tracks in the ECal takes up enormous amounts of time, so it is important to separate this function into its own processor, and adopt a new BDT. Additionally, we removed truth tracking of the electron with realistic tracking from our Recoil Tracker. We present the changes between the new BDT and its previous iteration, as well as its effectiveness in rejecting photonuclear and electronuclear backgrounds.