Uncertainty Propagation
Overview
Uncertainty propagation asks how uncertain inputs, parameters, and model predictions affect later calculations, forecasts, and decisions. Common approaches include linearization and the delta method, Monte Carlo sampling, Bayesian models, interval bounds, and learned probabilistic surrogates that carry distributions or covariances through complex systems. Applications range from scientific simulations and medical measurements to robotics and multi-stage AI systems, where benchmarks often assess calibration, coverage, sharpness, and downstream risk. A central challenge is balancing computational cost and useful precision while preserving dependencies: separately calibrated components can still produce unreliable results when their uncertainties interact.
Latest papers
44 papers4 papers in the last four weeks, up 33% on the four weeks before. 0.0% of all new papers.