Deep Learning in NOvA and DUNE

Chris Backhouse
(UCL)

NOvA and DUNE are two experiments looking to resolve the mysteries of the neutrino by studying the phenomenon of neutrino flavour change ("oscillation"). Their success is tied directly to their ability to resolve the flavour and properties of neutrino interactions observed in their detectors. Recent developments in Deep Learning have allowed NOvA to deploy a new classifier with substantially better performance than had been attained with traditional techniques.

I will give a brief overview of the two experiments and the history of machine learning, before describing the NOvA classifier, various extensions of the technique under development, and the promise similar techniques are currently showing in DUNE.