A Machine Learning Approach to Study the Electron-neutrino Charged-current Interaction On Iodine 127

An inclusive measurement of the cross section of the electron-neutrino charged-current interactions on $^{127}I$ will help study the quenching of $g_{A}$, the axial-vector coupling constant, which affects the rate of neutrinoless double beta decays. At the Los Alamos Meson Production Facility (LAMPF), an exclusive measurement was made but with a large statistical error. To make a first measurement of the inclusive cross section with low statistically uncertainty, a 185 kg NaI[Tl] prototype was deployed by the COHERENT collaboration. To reduce the major background, cosmic muons, a convolutional neural network (CNN) classifier and a decision tree classifier were developed. The best performer, tested with simulations, achieved a 95$\%$ classification accuracy assuming nano-second timing resolutions (77$\%$ accuracy otherwise). To address the non-linearity of NaI[Tl] crystals at high energies, calibrations using Michel electrons from stopped muon decays are underway.