Hi,

In the multinom function manual (nnet R-package) it is advised to "roughly scale" the variables to [0,1]. How necessary is that? Which is the best way (method) to do it?

Thanks in advance.

xavi.

Similar questions and discussions