October 2017
Beginner to intermediate
270 pages
7h
English
Classification models, as we have seen previously, take an image or other type of input as a parameter and return one array with as many elements as the number of the possible classes, with a corresponding probability for each one.
The normal architecture for this type of solution is a complex combination of convolutional and pooling layers with a logistic layer at the end, showing the probability any of the pretrained classes.
Read now
Unlock full access