Background rejection in NEXT using deep neural networks

We investigate the potential of using deep learning techniques to reject background events in searches for neutrinoless double beta decay with high pressure xenon time projection chambers capable of detailed track reconstruction. The differences in the topological signatures of background and signal...

Descripción completa

Detalles Bibliográficos
Autores: Renner, J.; Farbin, A.; Vidal, J.M.; Benlloch-Rodríguez, J.M.; Botas, A.; Ferrario, P.; Gómez-Cadenas, J.J.; Álvarez, V.; Azevedo, C.D.R.; Borges, F.I.G.; Cárcel, S.; Carrión, J.V.; Cebrián, S.; Cervera, A.; Conde, C.A.N.; Díaz, J.; Diesburg, M.; Esteve, R.; Fernandes, L.M.P.; Ferreira, A.L.; Freitas, E.D.C.; Goldschmidt, A.; González-Díaz, D.; Gutiérrez, R.M.; Hauptman, J.; Henriques, C.A.O.; Morata, J.A.H.; Herrero, V.; Jones, B.; Labarga, L.; Laing, A.; Lebrun, P.; Liubarsky, I.; López-March, N.; Lorca, D.; Losada, M.; Martín-Albo, J.; Martínez-Lema, G.; Martínez, A.; Monrabal, F.; Monteiro, C.M.B.; Mora, F.J.; Moutinho, L.M.; Nebot-Guinot, M.; Novella, P.; Nygren, D.; Palmeiro, B.; Para, A.; Pérez, J.; Querol, M.; Ripoll, L.; Rodríguez, J.; Santos, F.P.; Dos Santos, J.M.F.; Serra, L.; Shuman, D.; Simón, A.; Sofka, C.; Sorel, M.; Toledo, J.F.; Torrent, J.; Tsamalaidze, Z.; Veloso, J.F.C.A.; White, J.; Webb, R.; Yahlali, N.; Yepes-Ramírez, H.
Tipo de documento: Artículo
Estado:Published version
Fecha de publicación:2017
País:España
Institución:Universidad de Zaragoza
Repositorio:Zaguán. Repositorio Digital de la Universidad de Zaragoza
OAI Identifier:oai:zaguan.unizar.es:60704
Acceso en línea:http://zaguan.unizar.es/record/60704
Nivel de acceso:Acceso abierto
Descripción
Sumario:We investigate the potential of using deep learning techniques to reject background events in searches for neutrinoless double beta decay with high pressure xenon time projection chambers capable of detailed track reconstruction. The differences in the topological signatures of background and signal events can be learned by deep neural networks via training over many thousands of events. These networks can then be used to classify further events as signal or background, providing an additional background rejection factor at an acceptable loss of efficiency. The networks trained in this study performed better than previous methods developed based on the use of the same topological signatures by a factor of 1.2 to 1.6, and there is potential for further improvement.