Entering edit mode
6.7 years ago
Billy
•
0
Hi
If I have a n*m A-matrix (Ai,j) 1<i<n, 1<j<m,="" and="" i="" want="" to="" build="" a="" neural="" net="" (with="" neuralnetpackage="" in="" r)="" transforming="" columns="" (ai,k)="" in="" (ai,k+1).<="" p="">
I tried to make it with net<- neuralnet(formula =A[,k]~A[,k+1] , data=A, hidden=10), but there's this message
Error in model.frame.default(formula.reverse, data) : invalid type (list) for variable 'A[,k + 1]'
My goal is to build a neuralnet with m>>n to analyse scaling laws.
Can someone help me in this field?