An MLP consists of at least three layers of nodes: an input layer, a hidden layer and an output layer.
Also known as feedforward artificial neural network. An MLP consists of at least three layers of nodes: an input layer, a hidden layer and an output layer. Except for the input nodes, each node is a neuron that uses a nonlinear activation function. MLP utilizes a supervised learning technique called Backpropagation for training. It can distinguish data that is not linearly separable.