For questions related to the technique of backpropagation, whereby the loss, error, or correction signal calculated at the output of an artificial network output is fed back to the parameters in each layer of the network until the network's behavior converges to a training state within the required accuracy and reliability.

Backpropagation (or back propagation) is a method used in artificial networks to calculate a portion of the error contribution of each neuron after a training example, batch of them, or segment of a batch of them (mini-batch) is processed. It is a special case of successive approximation and searching, where the space to be searched is represented by a set of parameters that determine the behavior of the network. It is also related to gradient descent and the general technique of automatic differentiation.

history | excerpt history