Struct TverskyLoss 

pub struct TverskyLoss {
    pub alpha: f64,
    pub beta: f64,
    pub smooth: f64,
}

Tversky loss (generalization of Dice loss). Useful for handling class imbalance in segmentation.

Fields

alpha: f64

Alpha parameter (weight for false positives).

beta: f64

Beta parameter (weight for false negatives).

smooth: f64

Smoothing factor.

Trait Implementations

impl Clone for TverskyLoss

fn clone(&self) -> TverskyLoss

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for TverskyLoss

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for TverskyLoss

fn default() -> Self

Returns the “default value” for a type.

impl Loss for TverskyLoss

fn compute( &self, predictions: &ArrayView<'_, f64, Ix2>, targets: &ArrayView<'_, f64, Ix2>, ) -> TrainResult<f64>

Compute loss value.

fn gradient( &self, predictions: &ArrayView<'_, f64, Ix2>, targets: &ArrayView<'_, f64, Ix2>, ) -> TrainResult<Array<f64, Ix2>>

Compute loss gradient with respect to predictions.

fn name(&self) -> &str

Get the name of the loss function.