Losses¶
losses sub-module within the ketos.neural_networks module
This module provides loss functions
- Contents:
FScoreLoss class:
-
class
ketos.neural_networks.dev_utils.losses.
FScoreLoss
(beta=1.0, **kwargs)[source]¶ Bases:
tensorflow.python.keras.losses.Loss
Loss function based on the inverse of F-Score.
When instantiated, the resulting loss function expects the predictions in the ‘y_pred’ argument and the true labels in the ‘y_true’ argument.
- Args:
- beta:float
- The relative weight of recall in relation to precision.
- Examples:
If beta = 1.0, recall has same weight as precision If beta = 0.5, recall has half the weight of precision If beta = 2.0, recall has twice the weight of precision
-
call
(y_true, y_pred)[source]¶ Invokes the Loss instance.
- Args:
- y_true: Ground truth values. shape = [batch_size, d0, .. dN], except
sparse loss functions such as sparse categorical crossentropy where shape = [batch_size, d0, .. dN-1]
y_pred: The predicted values. shape = [batch_size, d0, .. dN]
- Returns:
Loss values with the shape [batch_size, d0, .. dN-1].