Back Propagation is a common method of training artificial neural networks so as to minimize objective function. This paper describes the implementation of back propagation algorithm.