kumulant

BoltzmannSpec

@Serializable
@SerialName(value = "Boltzmann")
data class BoltzmannSpec(val nbrArms: Int, val priorMean: Double = 0.0, val priorWeight: Double = 0.02, val initialTau: Double = 1.0, val minTau: Double = 0.001, val decay: Double = 1.0) : UnivariateBanditSpec(source)

Spec for BoltzmannBandit.

Constructors

BoltzmannSpec

constructor(nbrArms: Int, priorMean: Double = 0.0, priorWeight: Double = 0.02, initialTau: Double = 1.0, minTau: Double = 0.001, decay: Double = 1.0)(source)

Properties

decay

Cooling decay exponent: tau(t) = initialTau / t^decay.

initialTau

Initial softmax temperature.

minTau

Floor on the temperature schedule.

nbrArms

Number of arms in the population.

priorMean

Per-arm prior on the running reward mean.

priorWeight

Per-arm prior pseudo-count.