Multi-layer perceptrons

The multi-layer perceptron is a simple ANN. Its name, however, is a misnomer. A multi-layer perceptron is not a single perceptron with multiple layers, but rather multiple layers of artificial neurons that resemble perceptrons. Multi-layer perceptrons have three or more layers of artificial neurons that form a directed, acyclic graph. Generally, each layer is fully connected to the subsequent layer; the output, or activation, of each artificial neuron in a layer is an input to every artificial neuron in the next layer. Features are input through the Input layer. The simple neurons in the input layer are connected to at least one Hidden layer. Hidden layers represents latent variables; these cannot be observed in the training data. The hidden neurons in these layers are often called hidden units. Finally, the last hidden layer is connected to an Output layer; the activations of this layer are the predicted values of the response variable. The following diagram depicts...