Radial-Basis-Function-Network

A python implement of radial basis function network (RBFN), which can accept multi-variable input.

The dependencies:

• python3
• scikit-learn

Notes

1. How to specify the centers is the most difficult. I specify the centers with clustering centers. ( So you can use any clustering algorithm to get the centers. ). Now, you can specify the number of centers as whatever you want.
2. The scale length of each radial basis function can be different. But I set the scale length of each radial basis function with same value.

Test

Reference

• Broomhead, D.S. and Lowe, D., 1988. Radial basis functions, multi-variable functional interpolation and adaptive networks (No. RSRE-MEMO-4148). Royal Signals and Radar Establishment Malvern (United Kingdom).