tf.keras.metrics.CosineSimilarity

Computes the cosine similarity between the labels and predictions.

Inherits From: Mean, Metric, Layer, Module

cosine similarity = (a . b) / ||a|| ||b||

See: Cosine Similarity.

This metric keeps the average cosine similarity between predictions and labels over a stream of data.

name (Optional) string name of the metric instance.
dtype (Optional) data type of the metric result.
axis (Optional) Defaults to -1. The dimension along which the cosine similarity is computed.

Standalone usage:

# l2_norm(y_true) = [[0., 1.], [1./1.414], 1./1.414]]]
# l2_norm(y_pred) = [[1., 0.], [1./1.414], 1./1.414]]]
# l2_norm(y_true) . l2_norm(y_pred) = [[0., 0.], [0.5, 0.5]]
# result = mean(sum(l2_norm(y_true) . l2_norm(y_pred), axis=1))
#        = ((0. + 0.) +  (0.5 + 0.5)) / 2
m = tf.keras.metrics.CosineSimilarity(axis=1)
m.update_state([[0., 1.], [1., 1.]], [[1., 0.], [1., 1.]])
m.result().numpy()
0.49999997
m.reset_state()
m.update_state([[0., 1.], [1., 1.]], [[1., 0.], [1., 1.]],
               sample_weight=[0.3, 0.7])
m.result().numpy()
0.6999999

Usage with compile() API:

model.compile(
    optimizer='sgd',
    loss='mse',
    metrics=[tf.keras.metrics.CosineSimilarity(axis=1)])

Methods

reset_state

View source

Resets all of the metric state variables.

This function is called between epochs/steps, when a metric is evaluated during training.

result

View source

Computes and returns the metric value tensor.

Result computation is an idempotent operation that simply calculates the metric value using the state variables.

update_state

View source

Accumulates metric statistics.

y_true and y_pred should have the same shape.

Args
y_true Ground truth values. shape = [batch_size, d0, .. dN].
y_pred The predicted values. shape = [batch_size, d0, .. dN].
sample_weight Optional sample_weight acts as a coefficient for the metric. If a scalar is provided, then the metric is simply scaled by the given value. If sample_weight is a tensor of size [batch_size], then the metric 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 metric element of y_pred is scaled by the corresponding value of sample_weight. (Note on dN-1: all metric functions reduce by 1 dimension, usually the last axis (-1)).

Returns
Update op.