GradientDescentLinearRegression

public class GradientDescentLinearRegression : LinearRegression

Gradient Descent linear regression.

Reference: “linear regression using gradient descent”

  • The number of iterations for gradient descent.

    Declaration

    Swift

    var iterationCount: Int

  • The linear rate gradient descent.

    Declaration

    Swift

    var learningRate: Float

  • Whether to calculate the intercept for this model.

    Declaration

    Swift

    public var fitIntercept: Bool

  • The weights of the model.

    Declaration

    Swift

    public var weights: Tensor<Float>

  • Creates a linear regression model.

    Declaration

    Swift

    public init(
    iterationCount: Int = 1000,
    learningRate: Float = 0.001,
    fitIntercept: Bool = true
)

    Parameters

    iterationCount

    The number of iterations for gradient descent. The default is 1000.
    learningRate

    The learning rate for gradient descent. The default is 0.001.
    fitIntercept

    Whether to calculate the intercept for this model. If false, no intercept will be used in calculations. The default is true.

  • Initialize weights between [-1/N, 1/N].

    Declaration

    Swift

    internal func initializeWeights(featuresCount: Int)

    Parameters

    featuresCount

    The number of features in training data.

  • Fit a linear model.

    Declaration

    Swift

    public func fit(data: Tensor<Float>, labels: Tensor<Float>)

    Parameters

    data

    Training data with shape [sample count, feature count].
    labels

    Target value with shape [sample count, target count].