Artificial Neural Networks

Backpropagation

  1. Initialize all weights to small random numbers.
  2. For each training example Do
    1. Input the training example to the network and compute the network outputs
    2. For each output unit k calculate its error term δ k δ k o k (1-o k )(t k -o k )
    3. For each hidden unit h calculate its error term δ h δ h o h (1-o h ) k outputsw h ,kδ k
    4. Update each network weight w i ,j w i ,j w i ,j+Δw i ,j where Δ w i ,j=ηδ j x i ,j
  3. Goto 2 if termination condition is not met.

José M. Vidal .

21 of 33