Introduction - If you have any usage issues, please Google them yourself
The method gives the input weights and thresholds of neuron weights randomly, and then calculates the output weights by regularization principle. The neural network can still approximate any continuous system.