In this chapter, we've learned the basic concepts of multilayer neural networks and how to implement it in them Keras environment. To begin, the essential elements of an artificial neural network were analyzed. ANNs are mathematical models able to simulate the usual activities of the human brain. A neural network is composed of layers; neurons and connections between the neurons are arranged in different layers. Changing the weights of these connections determines the learning ability of a neural network.
Multilayer neural networks are neural networks composed of many hidden levels (at least two) organized hierarchically. Hierarchical organization allows you to share and reuse information. Along the hierarchy, you can select specific features and discard unnecessary details in order to maximize the invariance. Many variations of multilayer neural networks were introduced...
