The three main steps to defining a model:
Define the model structure
Configure the learning process
Train the model
There are two ways in which we can define our models using Keras: one is using sequential API and the other is using functional API
NeuralNets package can be used for simple models because it's not as fexible as Keras.
If we have a complex neural network architecture, functional API helps us in building the model; it allows us to use multiple layers/structures multiple times.
All the information of a model can be stored into a separate shareable and reusable file.
Callbacks help save models at each epoch.
Early stopping functionality should be used while creating a model, to stop overfitting the model
"Nós enviaremos as instruções para resetar a sua senha para o endereço associado. Por favor, digite o seu e-mail atual."