Backpropagation

Short for "backward propagation of errors." The algorithm that computes the gradient of the loss function with respect to every weight in a multi-layer network, by applying the chain rule backwards from the output layer to the input layer. The engine of all neural network training.