BP neural network learning machine demonstration
no vote
Application background Neural network is constructed by simulating human's neuron activity. Its basic unit is called "neuron", when the input is greater than a certain value, the neuron is excited, and the output is not responsive. And this input comes from all other neurons. The neuron's response function has a variety of (need to meet the micro, such a simple function can be fitted to any nonlinear function), select the SIGMOD function. The key of BP neural network is to deduce the formula of weight and threshold value.The basic idea of BP: the positive propagation of the signal is the error of the back propagation:- forward transmission: the input sample is passed from the input layer to the output layer. - back propagation: input error in some form through the hidden layer to the input layer by layer back propagation, and the error is apportioned to all the units of each layer, so as to obtain the error signal of the layers of units as