CMS-SUS-19-012
The AUC performance of the parametric neural networks for discriminating the signal from the total background predicted in simulation, as a function of δm for the trainings targeting different signal models. The top row corresponds to the neutral network targeting signals with WZ-mediated decays with different mass ranges to show all points. The bottom row corresponds to the models with slepton-mediated decays at x= 0.05 (left), x= 0.5 (middle) and x= 0.95 (right). Neural network models shown in blue are trained using all available δm points, those in red are trained with all available points except the point for which the performance is shown. The models in green are not parametric and only trained to find a signal at the point where the performance is indicated. Each neural network is retrained ten times, and the mean performances are shown, with error bars indicating the standard deviation computed from ten performance values. This means that each red and green point correspond to ten neural network trainings. The entire blue curve in each figure also corresponds to ten trainings.
Parametric neural network….