Jul 12 – 23, 2021
Europe/Berlin timezone

Bayesian Deep Learning for Shower Parameter Reconstruction in Water Cherenkov Detectors

Jul 13, 2021, 12:00 PM
1h 30m


Poster GAI | Gamma Ray Indirect Discussion


Clecio R. Bom (Centro Brasileiro de Pesquisas Físicas)


Deep Learning methods are among the state-of-art of several computer vision tasks, intelligent control systems, fast and reliable signal processing and inference in big data regimes. It is also a promising tool for scientific analysis such as gamma/hadron discrimination.
We present an approach based on Deep Learning for the regression of shower parameters, namely its core position and energy at the ground, using water Cherenkov detectors. We design our method using simulations. In this contribution, we explore the recovery of the shower’s center coordinates. We evaluate the limits of such estimation near the borders of the arrays, including the when the center is outside the detector’s range. We also address the feasibility of recovering other parameters, such as ground energy. We used Bayesian Neural Networks and derived and quantified systematic errors arising from Deep Learning models and optimized the network design. The method could be easily adapted to estimate other parameters.


Machine Learning, Deep Learning

Subcategory Experimental Methods & Instrumentation

Primary author

Clecio R. Bom (Centro Brasileiro de Pesquisas Físicas)


Ms Luciana O. Dias (Centro Brasileiro de Pesquisas Físicas) Ruben Conceição (LIP - Laboratório de Instrumentação e Física Experimental de Partículas) Bernardo Tomé (Laboratório de Instrumentação e Física Experimental de Partículas) Ulisses Barres de Almeida (Brazilian Center for Physics Research (CBPF)) Arthur Moraes (CBPF) Mário Pimenta (LIP/IST) Ronald Shellard Dr Márcio P. Albuquerque (Centro Brasileiro de Pesquisas Físicas)

Presentation materials