Loading
Uncertainty quantification attaches to each prediction an estimate of how much to trust it, so a system can report high confidence, low confidence or refuse.
Uncertainty quantification is the quantitative characterisation and estimation of uncertainties in computational and real-world applications (Wikipedia). For a clinical prediction the sources are whether the input is in distribution, how stable the prediction is under bootstrap refits, and how strong the underlying signal is; a shape that reports the prediction together with these gates lets a downstream reader (a clinician, or another program) act only on the confident ones. Conformal prediction is the version with a statistical guarantee.
Shares Bootstrap resampling, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Calibration: reliability diagrams and the Brier score, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Bootstrap resampling, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Local-only language models over patient data (privacy by architecture), Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Calibration: reliability diagrams and the Brier score, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Local-only language models over patient data (privacy by architecture), Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Out-of-distribution detection (Mahalanobis guard), Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Calibration: reliability diagrams and the Brier score, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.