from . import LossFunction
import numpy
[docs]class TangentialLoss(LossFunction):
"""Tangent loss function, as described in http://www.svcl.ucsd.edu/projects/LossDesign/TangentBoost.html."""
[docs] def loss(self, targets, scores):
"""The function computes the logit loss values using prediction scores and targets.
Keyword parameters:
targets (float <#samples, #outputs>): The target values that should be reached.
scores (float <#samples, #outputs>): The scores provided by the classifier.
Returns
(float <#samples, #outputs>): The loss values for the samples, always >= 0
"""
return (2. * numpy.arctan(targets * scores) - 1.) ** 2
[docs] def loss_gradient(self, targets, scores):
"""The function computes the gradient of the tangential loss function using prediction scores and targets.
Keyword parameters:
targets (float <#samples, #outputs>): The target values that should be reached.
scores (float <#samples, #outputs>): The scores provided by the classifier.
Returns
loss (float <#samples, #outputs>): The gradient of the loss based on the given scores and targets.
"""
m = targets * scores
numer = 4. * (2. * numpy.arctan(m) - 1.)
denom = 1. + m ** 2
return numer / denom