8-Neural Networks

What neural networks add to what we’ve seen previously is that we connect multiple such models together to obtain a network.

When talking about neural networks some terminology is slightly different: - Instead of coefficients we say weights - The non-linear part of the model, which in the case of logistic regression was the sigmoid function, is called the activation function - One such model is called a neuron - Neurons are connected to each other by letting the output of one neuron be an input of another.