Credible Prediction Model Based on BP Neural Network

Qiao Liu
2007-01-01
Abstract:The aim is to provide a kind of rule which is more accurate and more effective.This opitimization builds based on the traditional back propagation neural network model.Firstly,opitimize the traditional N-W rule for choosing the rudimentary power-number.Secondly,optimize other parameters.Advance the optimization tactics with using this model to forecast the system's dependable.Giving the forecast model depend on ANN'BP theroy,realise method and advantage.This way is feasible by the appraising example.
What problem does this paper attempt to address?