Machine Learning---neural network (2)

Today, we mainly study an important algorithm of neural network--BP (back propagation)

It is a feedforward network, that is, the output of the neuron is transmitted forward; its name means that the last output error is transmitted through the neural network and affects the weights of each level.

The simple idea is that the forward propagation of the signal----the reverse propagation of the > Error, and that the output error is passed through the hidden layer to the input layer in some form, and the weights are constantly updated.

This blog is well written: http://hahack.com/reading/ann2/

Machine Learning---neural network (2)