Mastering Machine Learning with scikit-learn - Second Edition

Feed-forward and feedback ANNs

ANNs are described by three components. The first is the model's architecture, or topology, which describes the types of neuron and the structure of the connections between them. Then we have the activation functions used by the artificial neurons. The third component is the learning algorithm that finds the optimal values of the weights.

There are two main types of ANN. Feed-forward neural networks are the most common type and are defined by their directed acyclic graphs. Information travels in one direction only, towards the output layer, in feed-forward neural networks. Conversely, feedback neural networks, or recurrent neural networks, contain cycles. The feedback cycles can represent an internal state for ...