newrb() RBF network - any preprocessing?

2 visualizaciones (últimos 30 días)
Frederick Turner
Frederick Turner el 4 de Feb. de 2018
Editada: Frederick Turner el 4 de Feb. de 2018
I'm using the newrb() function to generate radial basis function neural networks, and was expecting MATLAB to perform standardisation on the input and output data (i.e. using mapstd() to map the mean to zero and the variance to one for each feature); from what I understand of radial basis function networks, it is incredibly important to ensure that the variances of the variables are comparable in this way. However, for the networks I create, 'inputs{1}.processFcns' and 'inputs{1}.processSettings' are both empty arrays (even after configure/sim have been run). Does MATLAB perform any preprocessing, and if so, where can the preprocessing settings for the network be accessed?

Respuestas (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by