UncertaintyQuantificationOptions
Options & Configuration · AiDotNet.Models.Options
Configuration options for enabling uncertainty quantification during inference.
For Beginners
This lets you ask the model not only “what is the prediction?” but also “how sure are you?”
How It Works
Uncertainty quantification (UQ) augments standard point predictions with an uncertainty estimate. For supported model types, the library can sample multiple stochastic predictions and aggregate them into a mean prediction and an uncertainty estimate (variance).
Constructors
| Constructor | Summary |
|---|---|
UncertaintyQuantificationOptions() |
Properties
| Property | Summary |
|---|---|
ConformalPredictionMode ConformalMode { get; set; } |
Gets or sets the conformal calibration mode used when producing conformal intervals/sets. |
ProbabilityCalibrationMethod CalibrationMethod { get; set; } |
Gets or sets the probability calibration method used when calibration labels are provided. |
UncertaintyQuantificationMethod Method { get; set; } |
Gets or sets the uncertainty quantification strategy to use. |
bool DenormalizeUncertainty { get; set; } |
Gets or sets whether the returned uncertainty should be denormalized to match the output scale. |
bool EnableIsotonicRegressionCalibration { get; set; } |
Gets or sets whether to fit and apply isotonic regression calibration (binary calibration) when calibration labels are provided. |
bool EnablePlattScaling { get; set; } |
Gets or sets whether to fit and apply Platt scaling (binary calibration) when calibration labels are provided. |
bool EnableTemperatureScaling { get; set; } |
Gets or sets whether to fit and apply temperature scaling for classification-like outputs when calibration labels are provided. |
bool Enabled { get; set; } |
Gets or sets whether uncertainty quantification is enabled. |
double ConformalConfidenceLevel { get; set; } |
Gets or sets the desired conformal coverage level when using conformal prediction. |
double DeepEnsembleInitialNoiseStdDev { get; set; } |
Gets or sets the standard deviation of the initial parameter perturbation applied when constructing ensemble members. |
double LaplacePriorPrecision { get; set; } |
Gets or sets the prior precision (inverse variance) used by diagonal Laplace approximation. |
double MonteCarloDropoutRate { get; set; } |
Gets or sets the dropout rate used when injecting Monte Carlo Dropout layers automatically. |
double SwagLearningRate { get; set; } |
Gets or sets the learning rate used for SWAG posterior fitting on calibration data. |
int AdaptiveConformalBins { get; set; } |
Gets or sets the number of confidence bins used when ConformalMode is Adaptive. |
int CrossConformalFolds { get; set; } |
Gets or sets the number of folds used when ConformalMode is CrossConformal. |
int DeepEnsembleSize { get; set; } |
Gets or sets the number of independently trained models used for deep ensemble uncertainty estimation. |
int NumSamples { get; set; } |
Gets or sets the number of stochastic samples to draw when using sampling-based methods. |
int PosteriorFitMaxSamples { get; set; } |
Gets or sets the maximum number of samples used to fit Laplace/SWAG posteriors from calibration data. |
int SwagBurnInSteps { get; set; } |
Gets or sets the number of initial SWAG steps to skip before collecting snapshots. |
int SwagNumSnapshots { get; set; } |
Gets or sets the number of SWAG snapshots to collect when fitting a SWAG posterior. |
int SwagNumSteps { get; set; } |
Gets or sets the number of SWAG update steps used to collect snapshots. |
int? RandomSeed { get; set; } |
Gets or sets an optional random seed for reproducible Monte Carlo sampling. |