PointPosterior
@Serializable
@SerialName(value = "PointPosterior")
Point estimate plus optional isotropic Gaussian noise. SGD models have no posterior variance to draw from; exploration adds a constant std-dev shake on top of the point estimate.
Functions
evaluate
open override fun evaluate(snapshot: StochasticRegressionResult, x: VectorView, rng: Random, exploration: Double = 1.0): Double(source)
Closes to predict(x) + sd * ||x|| * N(0,1) since the per-coord noise terms are iid; one Gaussian draw instead of one per coordinate.
sample
open override fun sample(snapshot: StochasticRegressionResult, rng: Random, exploration: Double = 1.0): VectorView(source)
Draw a weight vector from the posterior at exploration variance scale. exploration = 0.0 collapses to the point estimate; 1.0 is the calibrated posterior.