|
The artificial neural network is an emerging interdisciplinary . It has been 50 years of history from 1943 . In the 50 years of development process , and constantly enrich the contents of the artificial neural network , more and more widely in the areas , in pattern recognition, system identification , intelligent control , image processing , fault diagnosis , disease diagnosis , economic forecasts the fields have been widely used. This paper, using the method of artificial neural network BP algorithm, to predict the actual economic system . While using artificial neural networks to regional economic forecast , in-depth study of the internal mechanism of the artificial neural network , to improve on the original model , and several more effective neural network algorithm . Specifically, this work focused on the following aspects : the feed-forward neural network model to summarize , the model proposed several effective improvement program . Including the model algorithm and the improvement of the network structure . Prior to the use of artificial neural network and BP algorithm , Liaoning Province, three industrial added value , gross domestic product ( GDP ) as well as several other economic indicators in the next few years was predicted . And the predicted results were analyzed , the economic development of Liaoning Province put forward some reasonable suggestions . Reasonable to determine the scale of the road network is an important content of the highway network planning , is an important indicator of macro - control of highway construction . In this paper, the artificial neural network model for the total mileage demand forecast for the road network in Henan Province , to come to a more reasonable and effective forecasting results . BP algorithm has some shortcomings , in order to improve the efficiency and stability of the multi-layer feedforward neural network learning , the paper summarizes several new neural network algorithm based on the variable metric method , conjugate gradient method and the method of least squares .
|