kumulant/com.eignex.kumulant.bandit.univariate/UCB1Normal

UCB1Normal

class UCB1Normal(val alpha: Double = 1.0, priorMean: Double = 0.0, priorWeight: Double = 0.02) : BanditPolicy<MomentsResult> (source)

UCB1-Normal (Auer et al. 2002). Variance-aware UCB for Gaussian rewards; uses the sample variance derived from the MomentsResult snapshot to scale the confidence bound. Reach for it when rewards are roughly Gaussian and unbounded; UCB1's [0, 1] assumption doesn't hold.

Forces exploration until each arm has at least 8 * ln(K) pulls (K is the arm count), then switches to the variance-aware score mean + alpha * sqrt(16 * variance * ln(K - 1) / armSamples).

Constructors

UCB1Normal

constructor(alpha: Double = 1.0, priorMean: Double = 0.0, priorWeight: Double = 0.02)

Properties

alpha

val alpha: Double

Exploration scale on the confidence-bound term.

arm

open override val arm: MomentsArm

Per-arm cumulator spec; determines the prior pseudo-counts, value encoding, and result shape that evaluate consumes.

Functions

addArm

open override fun addArm(snapshot: MomentsResult)

Hook called when a new arm joins the population. Lets stateful policies fold the new arm's snapshot into their global counters (UCB's total-samples, UCB1Normal's arm count). Default no-op.

createArm

open fun createArm(): SeriesStat<MomentsResult>

Allocate a fresh per-arm accumulator from the arm spec. Default delegates to arm.createStat(); override only if the policy needs a non-standard variant.

evaluate

open override fun evaluate(snapshot: MomentsResult, step: Long, rng: Random): Double

Score an arm given its current snapshot. Higher scores are preferred by the bandit. step is the global update count (for time-dependent exploration schedules); rng is the bandit's shared com.eignex.kumulant.bandit.Bandit.random (consumed by sampling policies).

removeArm

open override fun removeArm(snapshot: MomentsResult)

Hook called when an arm leaves the population. Inverse of addArm; lets stateful policies remove the departing arm's contribution from their global counters. Default no-op.

update

open fun update(stat: SeriesStat<MomentsResult>, value: Double, weight: Double = 1.0)

Fold an observed reward value (with optional weight) into the per-arm stat. Default applies arm.encode first; policies with global counters (UCB families) override to update their counter alongside the stat update.