A hybrid framework alternates Bayesian experimental design for physics parameters with gradient-based calibration of a neural network model-discrepancy term, gated by an ensemble Kalman information-gain indicator.
Turbulence modeling in the age of data
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Active Learning of Model Discrepancy with Bayesian Experimental Design
A hybrid framework alternates Bayesian experimental design for physics parameters with gradient-based calibration of a neural network model-discrepancy term, gated by an ensemble Kalman information-gain indicator.