Deep Learning Approaches for LHCb ECAL Reconstruction

The aim of the LHCb Upgrade II at the LHC is to operate at a luminosity of 1.5 x 1034 cm-2 s-1 to collect a data set of 300 fb-1. This will require a substantial modification of the current LHCb ECAL due to high radiation doses in the central region and increased particle densities. Advanced detector R&D for both new and ongoing experiments in HEP requires performing computationally intensive and detailed simulations as a part of the detector-design optimization process. We propose a versatile approach to this task that is based on machine learning and can substitute the most computationally intensive steps of the process while retaining the GEANT4 accuracy to details. The approach covers entire detector representation from the event generation to the evaluation of the physics performance. The approach allows us to use an arbitrary arrangement of calorimetric modules of different types, various signal and background conditions, tunable reconstruction algorithms, and desired physics performance metrics. Being combined with properties of detector and electronic prototypes obtained from beam tests, the approach becomes even more versatile. We focus on the Upgrade II of the LHCb ECAL under the requirements for operation under high luminosity conditions. We discuss the general design of the approach, and particular estimations including energy, timing and spatial resolution for the future LHCb ECAL setup under different pile-up conditions. This contribution presents an overview of the deep learning approaches that are proposed to be used for reconstruction of the LHCb ECAL at high luminosities.