Fast and precise model calculation for KATRIN using a neural network

TL;DR

This paper introduces a neural network-based approach to rapidly and accurately approximate the KATRIN experiment's physics model, significantly reducing computation time while maintaining high precision essential for neutrino mass measurements.

Contribution

The authors develop a neural network model that replicates the KATRIN beta-spectrum predictions with high accuracy, enabling faster computations for data analysis.

Findings

Achieves a three-order-of-magnitude speed-up in model calculations.

Maintains relative errors below 1e-4, meeting KATRIN's accuracy requirements.

Facilitates efficient analysis of large datasets in neutrino mass experiments.

Abstract

We present a fast and precise method to approximate the physics model of the Karlsruhe Tritium Neutrino (KATRIN) experiment using a neural network. KATRIN is designed to measure the effective electron anti-neutrino mass using the kinematics of beta-decay with a sensitivity of 200 meV at 90% confidence level. To achieve this goal, a highly accurate model prediction with relative errors below the 1e-4-level is required. Using the regular numerical model for the analysis of the final KATRIN dataset is computationally extremely costly or requires approximations to decrease the computation time. Our solution to reduce the computational requirements is to train a neural network to learn the predicted beta-spectrum and its dependence on all relevant input parameters. This results in a speed-up of the calculation by about three orders of magnitude, while meeting the stringent accuracy requirements of KATRIN.