Optimization of Neutrino Oscillation Parameters Using Differential Evolution

We show how the traditional grid based method for finding neutrino oscillation parameters Δm² and tan² θ can be combined with an optimization technique, Differential Evolution (DE), to get a significant decrease in computer processing time required to obtain minimal chi-square (χ²) in four different regions of the parameter space. We demonstrate efficiency for the two-neutrinos case. For this, the χ² function for neutrino oscillations is evaluated for grids with different density of points in standard allowed regions of the parameter space of Δm² and tan² θ using experimental and theoretical total event rates of chlorine (Homestake), Gallex+GNO, SAGE, Superkamiokande, and SNO detectors. We find that using DE in combination with the grid based method with small density of points can produce the results comparable with the one obtained using high density grid, in much lesser computation time.