The Backpropagation Neural Network was first described in 1974 by Paul Werbos . It however didn't gain recognition until 1986 through the works of David Rumelhart, Geoffrey E. Hinton and Ronald Williams. In summary the Backpropagation Neural Network is a training sample. It is a learning method supervised, and requiring the teacher to be able to facilitate or calculate the desired output, for a
. given input. Usually having one input, one output and one hidden layer, Backpropagation Networks are multilayer perceptrons. For more info visit: http://en.wikipedia.org/wiki/Multilayer_perceptron .
