Discrimination between neutrino events and backgrounds using pulse shape information in reactor neutrino experiments

Abstract

The application of Bayesian neural networks (BNN) to discriminate neutrino events from backgrounds in reactor neutrino experiments has been described in Xu et al. (2008) [1] and Xu et al. (2009) [2]. In the present paper, the pulse shape information for a fast signal of a neutrino event or a background event is used as a part of inputs to BNN to discriminate neutrino events from backgrounds. The numbers of photoelectrons received by PMTs and the delay time for a delayed signal are used as the other part of inputs to BNN (Xu et al., 2009) [2]. As a result, compared to Xu et al. (2009) [2], the identification efficiency of fast neutron background events is significantly improved using the BNN in the present paper. The other identification efficiencies are consistent with those in Xu et al. (2009) [2].