i need help for basics of Back propagation algorithm understanding. am a fresher in this so need that kind of help something like to start from the scratch.
I googled but ended up in some formulas n all in very dry theoretical explanation.can somebody suggest a way out.
am trying to write a code for BP in VHDL if any small piece of code are appreciated very much.
any books related for reference are also appreciated
(MODS sorry found other threads stating the same but none was answered so had to start with a new name)
Back propagation algorithm is very simple. Here you find the coefficients from the back (i.e. from the output side). That's the only way to find the coefficients of the blocks in the neural networks.
Classical Backpropagation routine is an iterative process wich can spend too much time for reach the result, dependig on amount of neurons. There are some new adaptative algorithms based on enhanced Backpropagation, wich propose to accelerate that convergence.