Abstract

Among the particles of the Standard Model, neutrinos are the least understood. Experiments worldwide are engaged in studying their properties and behavior, which could be linked to the matter-antimatter asymmetry of the Universe. Over the past several years, particle physicists have adapted techniques from the field of computer vision, whose tasks translate naturally for detector data analysis and simulation. Particle physics datasets are also a rich playground for new algorithmic approaches to data analysis. Neutrino experiments often look for rare signals in large amounts of data. Deep Learning techniques have yielded substantial improvements to the physics reach of many experiments already, and have redefined the limit to what is attainable in the realm of data collection, analysis, and R&D. Not only is neutrino physics benefiting from these techniques, but we are also contributing in new ways to algorithm design and utilization. This talk will discuss the main successes and newest developments of deep learning applications in the field of neutrino physics. The particulars of neutrino experiment data and tasks will be discussed, as well as lessons learned and future applications of machine learning to the field of neutrino physics.