Speaker
David Francois Droz
(Universite de Geneve (CH))
Description
DAMPE is a satellite-borne experiment in operations since December 2015, capable of detecting high energy cosmic rays and gamma rays. Its calorimeter, corresponding to about 31 radiation lengths allows to study electrons up to 10 TeV, making it the deepest calorimeter in space to date. A major complication at these energies is the discrimination between protons and electrons, as both particles leave similar detector signatures. We propose solving this problem using deep neural networks (DNNs), a fast emerging machine learning technique. DNNs can accurately discriminate between event classes using only basic detector variables, albeit at the cost of computation time. We present the technique and evaluate its performances using simulations.
Author
David Francois Droz
(Universite de Geneve (CH))
Co-authors
Dr
Stephan Zimmer
(Universite de Geneve (CH))
Xin Wu
(Universite de Geneve (CH))
