The scaled hyper-cube extended method for calculating the width of receptive field in Radial Basis Function Networks

The Radial Basis Function Networks are very useful for Machine Learning algorithms. The performance of the Radial Basis Function Network depends upon the width of the receptive field of a Radial Basis Function Neuron. This work presents a newer method for the calculation of the width of the receptive field of a Radial Basis Function Neuron. This approach solves the problem of curse of dimensionality.;The SHE Method for Calculating the Width of Receptive Fields in Radial Basis Function Networks uses rescaling of data if needed. A simpler rescaling algorithm has been presented in this research work. This rescaling algorithm has very little computational cost compared to other existing comparable algorithms. In fact, it is very efficient on larger datasets even.