Initializer

init(validationData:maxIterations:l1Penalty:l2Penalty:stepSize:convergenceThreshold:featureRescaling:)

Creates a new set of parameters.

Declaration

init(validationData: MLDataTable? = nil, maxIterations: Int = 10, l1Penalty: Double = 0, l2Penalty: Double = 0.01, stepSize: Double = 1.0, convergenceThreshold: Double = 0.01, featureRescaling: Bool = true)

Parameters

validationData

The dataset used to monitor how well the model is generalizing.

The default value is `nil` which will use an automatically sampled validation set.

maxIterations

The maximum number of passes through the data.

The default value is 10.

l1Penalty

Weight on the l1-regularizer. The l1Penalty zeros out small coefficients, indicating features that are not useful for the model.

The default value is 0 which prevents any values from being discarded.

l2Penalty

Weight of the l2-regularizer. The larger the l2Penalty the less variance in the model.

The default value is 0.01.

stepSize

The adjustment size that should be made by the underlying solver. Values close to 1.0 take an aggressive step based off feedback from each training iteration.

The default value is 1.0.

convergenceThreshold

The threshold with which to determine if the model has converged. Consider reducing this value for higher training accuracy, but beware of overfitting.

The default value is 0.01.

featureRescaling

Determines if the features should be preprocessed to ensure all features are on the same scale.

The default value is true.