A nonlinear feedback term is introduced into the evaluation equation of wei
ghts of the backpropagation algorithm for neural network, the network becom
es a chaotic one. For the purpose of that we can investigate how the differ
ent feedback terms affect the process of learning and forecasting, we use t
he model to forecast the nonlinear time series which is produced by Makey-G
lass equation. By selecting the suitable feedback term, the system can esca
pe from the local minima and converge to the global minimum or its approxim
ate solutions, and the forecasting results are better than those of backpro
pagation algorithm.