Hello all,
I am trying to use the neuralnet function from the neural package to train
neural networks. I am finishing my thesis over statistical learning in which I
am comparring the MSE values for various algorithms with random data sets, so I
do not have a specific application. I have two issues that I am looking for
help with.
1. When I train a network with a real data set, I ALWAYS get the mean value as
all of the predictions! I get beautiful results when I use a generated data
set, but only the mean for real data. Is there any explanation of this?
2. For some data sets I get this error:
"Error in neurons[[length.weights]] : attempt to select less than one
element"
I can not find anything about what this means.
I would be grateful for any help over either or both of these issues.
Thanks,
Donna
_________________________________________________________________
[[elided Hotmail spam]]
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
