Enum LossFunction 

pub enum LossFunction {
    SquaredLoss,
    AbsoluteLoss,
    HuberLoss,
    QuantileLoss,
    LogisticLoss,
    DevianceLoss,
    ExponentialLoss,
    ModifiedHuberLoss,
    PseudoHuber,
    Fair,
    LogCosh,
    EpsilonInsensitive,
    Tukey,
    Cauchy,
    Welsch,
}

Loss functions for gradient boosting

Variants

SquaredLoss

Least squares loss for regression

AbsoluteLoss

Absolute deviation loss for regression (robust)

HuberLoss

Huber loss for regression (robust)

QuantileLoss

Quantile loss for regression

LogisticLoss

Logistic loss for binary classification

DevianceLoss

Deviance loss for multiclass classification

ExponentialLoss

Exponential loss for AdaBoost

ModifiedHuberLoss

Modified Huber loss

PseudoHuber

Pseudo-Huber loss (robust)

Fair

Fair loss function (robust)

LogCosh

LogCosh loss (smooth approximation of Huber)

EpsilonInsensitive

Epsilon-insensitive loss

Tukey

Tukey’s biweight loss (robust)

Cauchy

Cauchy loss (robust)

Welsch

Welsch loss (robust)

Implementations

impl LossFunction

pub fn loss(&self, y_true: Float, y_pred: Float) -> Float

Compute loss value

pub fn gradient(&self, y_true: Float, y_pred: Float) -> Float

Compute gradient (negative derivative of loss w.r.t. prediction)

pub fn hessian(&self, y_true: Float, y_pred: Float) -> Float

Compute Hessian (second derivative of loss w.r.t. prediction)

pub fn is_robust(&self) -> bool

Check if the loss function is robust to outliers

Trait Implementations

impl Clone for LossFunction

fn clone(&self) -> LossFunction

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LossFunction

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

Formats the value using the given formatter.

impl PartialEq for LossFunction

fn eq(&self, other: &LossFunction) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for LossFunction

impl StructuralPartialEq for LossFunction