Multilayer perceptron is the most frequently used type of neural network. Its architecture is called feedforward, the signals in the network are transmitted in one direction, input to output. It is a supervised learning algorithm.
Learning takes place in the perceptron. Input data processed, the results are compared to the amount of error in the output and the changed weights. This supervised learning is carried through backpropagation.
A multilayer perceptron consists of different layers. The input layer transmits the input signals. Hidden layer is found between the input and output layers, it processes the signals sent by the input and there can be more than one hidden layer in a network. And the results after processing are found on the output layer. It generates decision areas and increases the computing ability compared to single layer networks.
Machine Learning:Multi Layer Perceptrons
Prof. Dr. Martin Riedmiller
Multilayer Perceptron and Neural Networks
Marius-Constantin Popescu, Valentina Balas, Liliana Perescu-Popescu and Nikos Mastoraki