Learning OpenCV 4 Computer Vision with Python 3 - Third Edition

Choosing the size of the output layer

For a classifier, the number of nodes in the output layer is, by definition, the number of classes the network can distinguish. Continuing with the preceding example of an animal classification network, we can use an output layer of four nodes if we know we are going to deal with the following animals: dog, condor, dolphin, and dragon(!). If we try to classify data for an animal that is not in one of these categories, the network will predict the class that is most likely to resemble this unrepresented animal.

Now, we come to a difficult problem – the size of the hidden layer.