Events in our system are self-managed.  Group and event managers are encouraged to review privacy and security settings, and adjust them if needed.  If you need assistance please contact Indico Support - contact Help at bottom of page.

October 31, 2018 to November 4, 2018
UBC Robson Square
America/Vancouver timezone

Energy Reconstruction in NOvA with Regression Convolutional Neural Networks

Nov 1, 2018, 6:09 PM
Room C215 (UBC Robson Square)

Room C215

UBC Robson Square


Prof. Jianming Bian (UC Irvine)


NOvA is a long-baseline accelerator neutrino oscillation experiment. It uses the upgraded NuMI beam from Fermilab and measures electron neutrino appearance and muon neutrino disappearance at its Far Detector in Ash River, Minnesota. NOvA aims to resolve the mass hierarchy, the CP violation and the octant of theta23. We will focus on the development and application of deep learning to the reconstruction tasks at NOvA. NOvA is a pioneer to use convolutional neural networks for event classification and particle identification in the neutrino community. Recently, we developed regression convolutional neural networks to estimate electron neutrino energy with direct raw detector pixel inputs. Compared with kinematics-based energy reconstruction, this method shows a significantly better energy resolution. The regression CNN also shows smaller systematic uncertainties from the simulation of neutrino interactions.

Primary author

Prof. Jianming Bian (UC Irvine)

Presentation materials