Enum LinearLoss

Namespace

AiDotNet.Models.Options

Assembly

AiDotNet.dll

Loss functions for linear classifiers.

public enum LinearLoss

Fields

Hinge = 0

Hinge loss - used for SVM-style max-margin classification.

Log = 2

Log loss - logistic regression style.

ModifiedHuber = 3

Modified Huber loss - robust to outliers.

Perceptron = 4

Perceptron loss - classic perceptron algorithm.

SquaredHinge = 1

Squared hinge loss - smoother version of hinge loss.