Statistical pattern recognition: application to νμ→ντ oscillation searches based on kinematic criteria
Metadatos
Mostrar el registro completo del ítemEditorial
Scuola Internazionale Superiore di Studi Avanzati (SISSA)
Materia
Neutrino Detectors Telescopes
Fecha
2004Referencia bibliográfica
Bueno, A.; et al. Statistical pattern recognition: application to νμ→ντ oscillation searches based on kinematic criteria. Journal of High Energy Physics, 11: 014 (2004). [http://hdl.handle.net/10481/29072]
Patrocinador
This work has been supported by the CICYT Grant FPA2002-01835. S.N. acknowledges support from the Ramon y Cajal Program.Resumen
Classic statistical techniques (like the multi-dimensional likelihood and the Fisher discriminant method) together with Multi-layer Perceptron and Learning Vector Quantization Neural Networks have been systematically used in order to find the best sensitivity when searching for νμ→ντ oscillations. We discovered that for a general direct ντ appearance search based on kinematic criteria: a) An optimal discrimination power is obtained using only three variables (Evisible, PmissT and ρl) and their correlations. Increasing the number of variables (or combinations of variables) only increases the complexity of the problem, but does not result in a sensible change of the expected sensitivity. b) The multi-layer perceptron approach offers the best performance. As an example to assert numerically those points, we have considered the problem of ντ appearance at the CNGS beam using a Liquid Argon TPC detector.