We have seen the characteristics of artificial neurons and the tasks performed by the activation functions. Now let's look more closely at the characteristics of NNs. NNs are made up of layers of neurons, which together form a network. NNs can also be interpreted as artificial neuron graphs in which a weight is associated with each connection.
We have said that by adding an adequate number of neurons to the NNs, it is possible to emulate the behavior of any continuous mathematical function. In practice, NNs are nothing but an alternative way of representing mathematical functions of arbitrary complexity. The power of NNs manifests itself in their ability to assemble the original features extracted ...
