package jsat.lossfunctions;
/**
* The AbsoluteLoss loss function for regression L(x, y) = |x-y|.
*
* This function is only one differentiable.
*
* @author Edward Raff
*/
public class AbsoluteLoss implements LossR
{
private static final long serialVersionUID = -3398199227407867808L;
/**
* Computes the absolute loss
*
* @param pred the predicted value
* @param y the target value
* @return the loss for the functions
*/
public static double loss(double pred, double y)
{
return Math.abs(y - pred);
}
/**
* Returns the derivative of the absolute loss
*
* @param pred the predicted value
* @param y the target value
* @return the derivative of the loss function
*/
public static double deriv(double pred, double y)
{
return Math.signum(pred - y);
}
public static double regress(double score)
{
return score;
}
@Override
public double getLoss(double pred, double y)
{
return loss(pred, y);
}
@Override
public double getDeriv(double pred, double y)
{
return deriv(pred, y);
}
@Override
public double getDeriv2(double pred, double y)
{
return 0;
}
@Override
public double getConjugate(double b, double pred, double y)
{
return b*y;
}
@Override
public double getDeriv2Max()
{
return 0;
}
@Override
public AbsoluteLoss clone()
{
return this;
}
@Override
public double getRegression(double score)
{
return score;
}
@Override
public double lipschitz()
{
//see Stochastic Dual Coordinate Ascent Methods for Regularized Loss Minimization
//"absolute deviation loss are 1-Lipschitz."
return 1.0;
}
}
Ads help maintain this website.