Neutrinos are elusive particles, a bit apart in the Standard Model of physics. The discovery of their massive nature, via the observation of their oscillations more than 25 years ago, opened the door to a new field of study. Tremendous progress has been made so far in understanding the physics of neutrinos, and at present, we are entering an era of precision measurements for the parameters that describe their oscillations. The next long-baseline neutrino accelerator experiments will be crucial for these precision measurements and will need to feature innovative detection technologies that can offer high target mass, while also reducing various systematic uncertainties and providing high energy resolutions.
Getting out the most of the physics potential of these new, more capable detectors will be challenging in many ways: a large quantity of data will have to be analyzed, reconstruction of complex event topologies will be required, and intricate calibrations and simulation tunings will be necessary. AI and Machine Learning techniques will likely be key in addressing these challenges as demonstrated both by their remarkable success in many other fields, and early applications in neutrino physics for tasks such as particle reconstruction or differentiable simulation. Combining together all these recent developments might ultimately allow for the design of a fully ML-based data-driven analysis chain.
https://stanford.zoom.us/j/98973156241?pwd=cEU5RFdlVXoyc0JTeTlDMkozKzQ5UT09
Jamie Ryan, Zhi Zheng (jlryan@slac, zzheng@slac)
