Implementación computacional basada en el método estadístico Q-factors para separar señal de ruido en mediciones físicas

Abstract

In this work we present the statistical and computational considerations for the application of the statistical method called Q-factor to problems in physical measurements. This method allows the signal to be separated from noise in a data distribution, for this a multivariate analysis is done, with which it can be assigned a statistical weight to each event or individual within the measured population. To validate the optimal functioning of the codes developed in C++ language and supported by the data analysis framework ROOT, two applications are made: the first one to the field of high energy physics and the other to extragalactic astronomy, which were simulated with the Montecarlo method. The results that we obtained show an optimal implementation since the statistical values of the Montecarlo distributions contrasted with those obtained by the Q-factors method, show minimal differences.