School of Computing. Dublin City University. Home Blog Teaching Research Contact My big idea: Ancient Brain 


4 input nodes, 2 hidden, 4 output.
Exemplars:
Input 1 0 0 0 Output 1 0 0 0 Input 0 1 0 0 Output 0 1 0 0 Input 0 0 1 0 Output 0 0 1 0 Input 0 0 0 1 Output 0 0 0 1i.e. We simply want it to represent these 4 inputs somehow in just 2 hidden nodes.
Make the hidden nodes produce simply 0 or 1 output.
Then Question:
After training:
For each exemplar,
what do the 2 hidden outputs y_{j} look like?