BP neural network in Matlab.
The program is used to generate and train a BP neural network with the primary application of Matlab. I choose Sigmoid as the activation function of the hidden layer, and Linear function as the activation function of the output layer.
Before running the program, you should add all the files into Matlab path. It is recommended to run the program in BPtrain.m unless you want to improve my code.
You can change the training set, number of neurons in hidden layer, learning rate, repeat times and check interval in BPtrain.m by simply changing the default values.
If your input size and output size are both one, you can run film.m to generate a film showing the training process. An example is given below.
Enjoy your time with BPNN and Matlab! Feel free to raise an issue!
Created 7/9/2018 by Bill in Beijing.
Last modified 15/8/2021 by Bill in Chengdu.