@MASTERSTHESIS{ 2023:1217673175,
 	title = {Supervised neural network approach to modeling DUNE's LArTPC photon detector device},
 	year = {2023},
 	url = "https://bdtd.unifal-mg.edu.br:8443/handle/tede/2416",
 	abstract = "This work presents the development and evaluation of an artificial neural network (ANN)
 as a supervised learning model to complement the ArapucaSim software simulation of
 the behavior of Arapucas’ photon absorption probabilities on DUNE. DUNE will be a
 neutrino detector intended to address fundamental questions about the nature of elementary particles and their role in the universe. The Arapucas are the light-trapping devices
 proposed for the detection system of the DUNE far detector. A neural network model
 employed is a regressor model that receives as inputs the coordinates of a photon generator, along with positions on the surface device where the photon collides and produces
 outputs consisting in absorption probability for each photon. The input data is obtained
 from Geant4 simulations, specifically from the ArapucaSim module. Two cases were studied: one in which the photons arrive with normal incidence in the Arapuca surface, and
 the other including the angle dependence. While the Geant4 approach required hours to
 generate results, after training the neural network model, comparable probabilities are
 produced in seconds with high accuracy. This work shows the potential of neural network
 models as efficient alternatives for simulations in predicting photon absorption probabilities by the light sensor with less time and computational effort.",
 	publisher = {Universidade Federal de Alfenas},
 	scholl = {Programa de Pós-graduação em Física},
 	note = {Instituto de Ciência e Tecnologia}
}