Título: Um Classificador Neural On-line para a Física de Partículas
Autores: Damazio, Denis O.; Silva, Paulo V. M.; Seixas, José M.
Resumo: An on-line neural system is being developed for detecting outsiders in experimental high-energy particle beams. This detection is based on exploring the information provided by a calorimeter, which is one of the main detectors used in modern particle collider experiments, as it measures the energy of the incoming particles. The methodology for training the network with incoming information is described and the discrimination of outsider pions, muons and electrons are achieved with an efficiency better than 93%, from a simulation of the on-line system. The high quality of the neural detection is attested from a classical off-line analysis often used by experts in calorimetry. Implementation of the system on digital signal processor technology is also evaluated.
Código DOI: 10.21528/CBRN2001-044
Artigo em pdf: 5cbrn_044.pdf
Arquivo BibTex: 5cbrn_044.bib