Backpropagation

In machine learning, backpropagation is a widely used algorithm for training feedforward artificial neural networks or other parameterized networks with differentiable nodes. It is an efficient application of the Leibniz chain rule to such networks. It is also known as