Identification of the quark jet charge in e +e −→W +W − using neural networks and discriminant analysis methods

Abstract

Using multivariate analysis methods with some momentum weighted charge variables, an attempt is made to separate between jets coming from W's decays in the reactions e +e −→W +W −. This analysis is confined to the jets extracted from the generated Lund Monte Carlo events. Neural network performs better than discriminant analysis. Both methods were correct in more than 85% for classifying W + jets and W − jets.