October 2023
Beginner
192 pages
5h 40m
English

Classical machine learning models struggle with appropriate feature selection, feature vector dimensionality, and the inability to learn from the structure inherent in the input. Convolutional neural networks (CNNs) overcome these issues by learning to generate new representations of their inputs while simultaneously classifying them, a process known as end-to-end learning. CNNs are the representation-learning data processors I referred to in Chapter 2.
Elements of what became CNNs appeared at various times throughout the history of neural networks, beginning with Rosenblatt’s Perceptron, but the ...
Read now
Unlock full access