View source on GitHub |
Converts a class vector (integers) to binary class matrix.
tf.keras.utils.to_categorical(
y, num_classes=None, dtype='float32'
)
E.g. for use with categorical_crossentropy
.
Returns | |
---|---|
A binary matrix representation of the input as a NumPy array. The class axis is placed last. |
Example:
a = tf.keras.utils.to_categorical([0, 1, 2, 3], num_classes=4)
print(a)
[[1. 0. 0. 0.]
[0. 1. 0. 0.]
[0. 0. 1. 0.]
[0. 0. 0. 1.]]
b = tf.constant([.9, .04, .03, .03,
.3, .45, .15, .13,
.04, .01, .94, .05,
.12, .21, .5, .17],
shape=[4, 4])
loss = tf.keras.backend.categorical_crossentropy(a, b)
print(np.around(loss, 5))
[0.10536 0.82807 0.1011 1.77196]
loss = tf.keras.backend.categorical_crossentropy(a, a)
print(np.around(loss, 5))
[0. 0. 0. 0.]