pulumi-azure-native-kotlin/com.pulumi.azurenative.machinelearningservices.kotlin.inputs/StackEnsembleSettingsArgs

StackEnsembleSettingsArgs

data class StackEnsembleSettingsArgs(val stackMetaLearnerKWargs: Output<Any>? = null, val stackMetaLearnerTrainPercentage: Output<Double>? = null, val stackMetaLearnerType: Output<Either<String, StackMetaLearnerType>>? = null) : ConvertibleToJava<StackEnsembleSettingsArgs>

Advances setting to customize StackEnsemble run.

Constructors

Link copied to clipboard
fun StackEnsembleSettingsArgs(stackMetaLearnerKWargs: Output<Any>? = null, stackMetaLearnerTrainPercentage: Output<Double>? = null, stackMetaLearnerType: Output<Either<String, StackMetaLearnerType>>? = null)

Functions

Link copied to clipboard
open override fun toJava(): StackEnsembleSettingsArgs

Properties

Link copied to clipboard
val stackMetaLearnerKWargs: Output<Any>? = null

Optional parameters to pass to the initializer of the meta-learner.

Link copied to clipboard
val stackMetaLearnerTrainPercentage: Output<Double>? = null

Specifies the proportion of the training set (when choosing train and validation type of training) to be reserved for training the meta-learner. Default value is 0.2.

Link copied to clipboard
val stackMetaLearnerType: Output<Either<String, StackMetaLearnerType>>? = null

The meta-learner is a model trained on the output of the individual heterogeneous models.