Speaker
Daniel Li
(Brown University (US))
Description
We investigate modern machine learning techniques to derive calibration for the combined CMS electromagnetic and hadronic calorimeter system. We use the dataset from a 2006 CMS test beam to measure the calorimeter responses to pion beams of various energies. The performance of the network is evaluated by studying the linearity of calibrated responses. A convolutional neural network approach is used to train on a range of beam momenta from $2$ to $200\ GeV/c$ and to apply the correction to the energy distribution.
Primary author
Daniel Li
(Brown University (US))
Co-authors
Sergei Gleyzer
(University of Florida (US))
Meenakshi Narain
(Brown University (US))
Ulrich Heintz
(Brown University (US))
Sitong An
(CERN, Carnegie Mellon University (US))
Jason Terry
(Brown University)
Andrew Dabydeen
(Brown University)
Emanuele Usai
(Brown University (US))
