Help protect the Great Barrier Reef with TensorFlow on Kaggle Join Challenge


Implements the Weighted Kappa loss function.

Weighted Kappa loss was introduced in the Weighted kappa loss function for multi-class classification of ordinal data in deep learning. Weighted Kappa is widely used in Ordinal Classification Problems. The loss value lies in $ [-\infty, \log 2] $, where $ \log 2 $ means the random prediction.


kappa_loss = tfa.losses.WeightedKappaLoss(num_classes=4)
y_true = tf.constant([[0, 0, 1, 0], [0, 1, 0, 0],
                 [1, 0, 0, 0], [0, 0, 0, 1]])
y_pred = tf.constant([[0.1, 0.2, 0.6, 0.1], [0.1, 0.5, 0.3, 0.1],
                 [0.8, 0.05, 0.05, 0.1], [0.01, 0.09, 0.1, 0.8]])
loss = kappa_loss(y_true, y_pred)
<tf.Tensor: shape=(), dtype=float32, numpy=-1.1611925>

Usage with tf.keras API:

model = tf.keras.Model()
model.compile('sgd', loss=tfa.losses.WeightedKappaLoss(num_classes=4))

<... outputs should be softmax results if you want to weight the samples, just multiply the outputs by the sample weight ...>

num_classes Number of unique classes in your dataset.
weightage (Optional) Weighting to be considered for calculating kappa statistics. A valid value is one of ['linear', 'quadratic']. Defaults to 'quadratic'.
name (Optional) String name of the metric instance.
epsilon (Optional) increment to avoid log zero, so the loss will be $ \log(1 - k + \epsilon) $, where $ k $ lies in $ [-1, 1] $. Defaults to 1e-6.

ValueError If the value passed for weightage is invalid i.e. not any one of ['linear', 'quadratic']



Instantiates a Loss from its config (output of get_config()).

config Output of get_config().

A Loss instance.


View source

Returns the config dictionary for a Loss instance.


Invokes the Loss instance.

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]
sample_weight Optional sample_weight acts as a coefficient for the loss. If a scalar is provided, then the loss is simply scaled by the given value. If sample_weight is a tensor of size [batch_size], then the total loss for each sample of the batch is rescaled by the corresponding element in the sample_weight vector. If the shape of sample_weight is [batch_size, d0, .. dN-1] (or can be broadcasted to this shape), then each loss element of y_pred is scaled by the corresponding value of sample_weight. (Note ondN-1: all loss functions reduce by 1 dimension, usually axis=-1.)

Weighted loss float Tensor. If reduction is NONE, this has shape [batch_size, d0, .. dN-1]; otherwise, it is scalar. (Note dN-1 because all loss functions reduce by 1 dimension, usually axis=-1.)

ValueError If the shape of sample_weight is invalid.