Numerical analysis of neutrino physics within a high scale supersymmetry model via machine learning

A machine learning method is applied to analyze lepton mass matrices numerically. The matrices were obtained within a framework of high scale SUSY and a flavor symmetry, which are too complicated to be solved analytically. In this numerical calculation, the heuristic method in machine learning is adopted. Neutrino masses, mixings, and CP violation are obtained. It is found that neutrinos are normally ordered and the favorable effective Majorana mass is about $7\times {10}^{-3}$ eV.