I am also getting this error. Has anyone found a reason? My net has the following layers:
Input: Linear
delay: Delay
Hidden: Sigmoid
Output: SoftMax
FullSynapse's between all layers.
I am trying to do a 1 of 3 or 1 of 5 classification model that has very noisy input. If I set my output to any of the other types (Sigmoid, Tanh, etc) it works, but with multiple outputs it really...
2009-06-12 19:05:49 UTC in Java Object Oriented Neural Engine
