Neutrino Event Reconstruction with Machine Learning on NOvA

The NOvA experiment has detected the disappearance of muon (anti-)neutrinos and the appearance of electron (anti-)neutrinos in the NuMI beam at Fermilab and have made measurements of parameters related to neutrino oscillations including the neutrino mass hierarchy and the CP violating phase. Key to these measurements is the identification of neutrino events and the reconstruction of their energies, for which NOvA has developed some of the first machine learning implementations in the field. Further applications of machine learning are possible using new algorithms in semantic segmentation, a technique for classification of individual elements of an image, which are also applicable to neutrino events. I will present an application of machine learning for doing full neutrino event reconstruction utilizing instance aware semantic segmentation.