Library: | rpclib |
See also: | xgobilocpol xgobilowess xgobinadawat |
Quantlet: | xgobineunet | |
Description: | conmputes a neural network fit for XGobi |
Usage: | yh = xgobineunet(x,y,g,hidden) | |
Input: | ||
x | n x 1 matrix | |
y | n x 1 matrix | |
g | n x 1 matrix with integers starting at zero | |
hidden | scalar number of hidden units | |
Output: | ||
yh | n x 1 matrix smoothed fit |
; load the library for XGobi library ("rpclib") ; create a display d=createdisplay(1,1) ; read a data set x=read("motcyc") ; generate randomly three subgroups g=floor(3.*uniform(rows(x))) ; color the data after the subgroups setmaskp (x, g, 3, 4) ; show the data and the smooth show (d, 1, 1, x, x[,1]~xgobineunet(x[,1],x[,2],g,10))
shows a neural network fit of for the three different subgroups in the data
Library: | rpclib |
See also: | xgobilocpol xgobilowess xgobinadawat |