Speaker
Mr
Lorenz Vogel
(Heidelberg University)
Description
Abstract: Utilizing modern ML-techniques, we address the challenge of multi-dimensional correlated calibration of topological calorimeter-cell clusters (topo-clusters). Our Bayesian neural network (BNN) approach not only yields a continuous, unbinned calibration function that improves performance relative to the standard calibration but also provides single-cluster uncertainties. A boosted training of the BNN further improves the uncertainty estimate and the network precision in critical phase-space regions.
