I have a couple of questions on how to code the backpropagation algorithm of neural networks:
The topology of my networks is the input layer, the hidden layer, and the output layer. Both the hidden layer and the output layer have sigmoid functions.
- First of all, should I use bias? Where should I connect the offset to my network? Should I put one offset unit per layer in both the hidden and the output layer? What about the input layer?
- In this link, they define the last delta as I / O and propagate the deltas back as shown. They hold a table to place all deltas until the errors actually propagate in the forward direction. Is this a departure from the standard backpropagation algorithm?
- ?
- - , Resilient Propagation ?
: . d f1 (e)/de, , f1 (e) * [1- f1 (e)], ? 