PlattCalibratorResult

@Serializable

@SerialName(value = "PlattCalibratorResult")

data class PlattCalibratorResult(val slope: Double, val intercept: Double, val totalWeights: Double) : HasObservationCount(source)

Snapshot from PlattCalibratorStat: the learned sigmoid parameters and a helper that maps a raw classifier score to a calibrated probability.

The calibration map is 1 / (1 + exp(-(slope * x + intercept))).

Constructors

constructor(slope: Double, intercept: Double, totalWeights: Double)(source)

Sigmoid intercept.

Sigmoid slope; positive when higher raw scores correlate with positive labels.

Cumulative observation weight folded in.

True when no observation has been folded in, so every other field is a placeholder.

Calibrated probability for raw score x via sigmoid(slope * x + intercept).