Source code for slugnet.loss

import numpy as np


class Objective(object):
    """An objective function (or loss function, or optimization score
    function) is one of the two parameters required to compile a model.

    """
    def forward(self, outputs, targets):
        raise NotImplementedError()

    def backward(self, outputs, targets):
        """Backward function.

        Parameters
        ----------
        outputs, targets : numpy.array
            The arrays to compute the derivatives of them.

        Returns
        -------
        numpy.array
            An array of derivative.
        """
        raise NotImplementedError()

    def __str__(self):
        return self.__class__.__name__


class BinaryCrossEntropy(Objective):
    r"""
    Standard binary cross-entropy loss function.

    Binary cross-entropy is given by

    .. math::

        \bm{\ell}(\bm{\hat{y}}, \bm{y}) = - \frac{1}{N} \sum_{i=1}^N
            [\bm{y}_i \, \text{log}(\bm{\hat{y}}_i) + (1 - \bm{y}_i) \text{log}(1 - \bm{\hat{y}}_i)]

    """
    def __init__(self, epsilon=1e-11):
        self.epsilon = epsilon

    def forward(self, yh, y):
        yh = np.clip(yh, self.epsilon, 1 - self.epsilon)
        loss = -np.sum(y * np.log(yh) + (1 - y) * np.log(1 - yh), axis=1)
        return np.mean(loss)

    def backward(self, outputs, targets):
        outputs = np.clip(outputs, self.epsilon, 1 - self.epsilon)
        divisor = np.maximum(outputs * (1 - outputs), self.epsilon)

        return (outputs - targets) / divisor


class SoftmaxCategoricalCrossEntropy(Objective):
    def __init__(self, epsilon=1e-11):
        self.epsilon = epsilon

    def forward(self, outputs, targets):
        outputs = np.clip(outputs, self.epsilon, 1 - self.epsilon)
        return np.mean(-np.sum(targets * np.log(outputs), axis=1))

    def backward(self, outputs, targets):
        outputs = np.clip(outputs, self.epsilon, 1 - self.epsilon)
        return outputs - targets