Microsoft.ML.Calibrators Namespace

Namespace containing components for binary classification calibration.

Classes

CalibratedModelParametersBase	Base class for calibrated model parameters.
CalibratedModelParametersBase<TSubModel,TCalibrator>	Class for allowing a post-processing step, defined by Calibrator, to SubModel's output.
CalibratorEstimatorBase<TICalibrator>	Base class for calibrator estimators.
CalibratorTransformer<TICalibrator>	An instance of this class is the result of calling Fit(IDataView). If you pass a scored data, to the CalibratorTransformer<TICalibrator> Transform method, it will add the Probability column to the dataset. The Probability column is the value of the Score normalized to be a valid probability. The CalibratorTransformer<TICalibrator> is an instance of ISingleFeaturePredictionTransformer<TModel> where score can be viewed as a feature while probability is treated as the label.
FixedPlattCalibratorEstimator	Obtains the probability values by applying the sigmoid: f(x) = 1 / (1 + exp(-slope * x + offset). Note that unlike, say, PlattCalibratorEstimator, the fit function here is trivial and just "fits" a calibrator with the provided parameters.
IsotonicCalibrator	The isotonic calibrator.
IsotonicCalibratorEstimator	The isotonic calbrated estimator.
IsotonicCalibratorTransformer	The ITransformer implementation obtained by training a IsotonicCalibratorEstimator
NaiveCalibrator	The naive binning-based calibrator.
NaiveCalibratorEstimator	The naive binning-based calibrator estimator.
NaiveCalibratorTransformer	The ITransformer implementation obtained by training a NaiveCalibratorEstimator
PlattCalibrator	The Platt calibrator calculates the probability following: P(x) = 1 / (1 + exp(Slope * x + Offset)
PlattCalibratorEstimator	The Platt calibrator estimator.
PlattCalibratorTransformer	The ITransformer implementation obtained by training a FixedPlattCalibratorEstimator or a PlattCalibratorEstimator.

Interfaces

ICalibrator

An interface for probability calibrators.

Remarks

These components calculate probabilities from scores, for binary classifiers.