Skip to content

Latest commit

 

History

History
4 lines (3 loc) · 450 Bytes

README.md

File metadata and controls

4 lines (3 loc) · 450 Bytes

RBFNN_Demo_Fitting_Any_Function

A simple radial basis function neural network's demo, based on the matlab. It can fit any function only if it has a series of suitable parameters including the Gaussian kernel‘s number and range, learning rate and the parameter b.

If you really confuse about the RBF's internal principles, and afraid of coding, wish it will help you.Any question, you can contact me with the email 3527623856@qq.com. Having fun.