Speaker
Henning Bahl
Description
Ultra-fast, precise, and controlled amplitude surrogates are essential for future LHC event generation. First, we investigate the noise reduction and biases of network ensembles and outline a new method to learn well-calibrated systematic uncertainties for them. We also establish evidential regression as a sampling-free method for uncertainty quantification. In a second part, we tackle localized disturbances for amplitude regression and demonstrate that learned uncertainties from Bayesian networks, ensembles, and evidential regression all identify numerical noise or gaps in the training data.
Authors
Henning Bahl
Nina Elmer
Ramon Winterhalder
(University of Milan)
Tilman Plehn
(Institut für Theoretische Physik, Universität Heidelberg; Interdisciplinary Center for Scientific Computing (IWR), Universität Heidelberg)