A modular classification system based on neural networks, for the particle identification task in a physics experiment, is proposed. The system is oriented to prevent systematic classification errors that could occur analysing experimental data with different statistical feature distribution in respect of the expected ones. The system has been investigated in the context of the positron/proton classification problem in a cosmic ray space experiment, where the physics detectors are faced by the critical flight conditions. Finally, the experimental results shows as an adaptive training based on the real data can improve the classification model of the system.
Cooperative neural system for particle classification in a cosmic ray space experiment / Bellotti, R; Castellano, M; Demarzo, C; Pasquariello, G; Satalino, G. - In: INTERNATIONAL JOURNAL OF NEURAL SYSTEMS. - ISSN 0129-0657. - STAMPA. - 6:(1995), pp. 275-279. (Intervento presentato al convegno 3rd Workshop Neural Networks : From Biology to High Energy Physics tenutosi a Isola d'Elba, Italy nel September 26-30, 1994).
Cooperative neural system for particle classification in a cosmic ray space experiment
Castellano, M;
1995-01-01
Abstract
A modular classification system based on neural networks, for the particle identification task in a physics experiment, is proposed. The system is oriented to prevent systematic classification errors that could occur analysing experimental data with different statistical feature distribution in respect of the expected ones. The system has been investigated in the context of the positron/proton classification problem in a cosmic ray space experiment, where the physics detectors are faced by the critical flight conditions. Finally, the experimental results shows as an adaptive training based on the real data can improve the classification model of the system.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
