Accord.Math.Optimization.Losses Namespace

Absolute loss, also known as L1-loss.

Accuracy loss, also known as zero-one-loss. This class provides exactly the same functionality as ZeroOneLoss but has a more intuitive name. Both classes are interchangeable.

Binary cross-entropy loss for multi-label problems, also known as logistic loss per output of a multi-label classifier.

Categorical cross-entropy loss for multi-class problems, also known as the logistic loss for softmax (categorical) outputs.

Euclidean loss, also known as zero-one-loss. This class provides exactly the same functionality as SquareLoss but has a more intuitive name. Both classes are interchangeable.

Mean Accuracy loss, also known as zero-one-loss per class. Equivalent to ZeroOneLoss but for multi-label classifiers.

Negative log-likelihood loss.

Base class for loss functions.

Base class for loss functions.

R² (r-squared) loss.

Square loss, also known as L2-loss or Euclidean loss.

Accuracy loss, also known as zero-one-loss.

Hinge loss.

Logistic loss.

Smooth Hinge loss.

Squared Hinge loss.

Common interface for differentiable loss functions, such as SquareLoss, HingeLoss, CategoryCrossEntropyLoss and BinaryCrossEntropyLoss.

Common interface for loss functions, such as SquareLoss, HingeLoss and CategoryCrossEntropyLoss.

Common interface for loss functions, such as SquareLoss, HingeLoss, CategoryCrossEntropyLoss and BinaryCrossEntropyLoss.