Data Analysis of Particle Physics Experiments Based on Machine Learning and the Mitchell’s Criteria

Abstract

Commonly the searching and identification of new particles, requires to reach highest efficiencies and purities as well. It demands to apply a chain of cuts that reject the background substantially. In most cases the processes to extract signal from the background is carried out by hand with some assistance of well designed and intelligent codes that save time and resources in high energy physics experiments. In this paper we present one application of the Mitchell’s criteria to extract efficiently beyond Standard Model signal events yielding an error of order of 1.22%. The usage of Machine Learning schemes appears to be advantageous when large volumes of data need to be scrutinized.