The system of economic forecasts is a highly uncertain nonlinear system. It is affected by a number of factors ,which have complicated correlation . So, the traditional forecasting methods about economic forecasts have a lot of Shortcomings. But BP neural network can overcome these constraints and achieve precise forecast nonlinear.This paper firstly uses the principal components analysis to abstract the information, then trains the net by the algorithm based on Bayesian regulation. The algorithm based on Bayesian regulation is better than the standard BP algorithm, by changing its objective function and using Bayesian methods to select parameters automatically, thereby enhancing generalization. This paper Analysis the data on the level of economic development from 1985 to 2005 on Hubei Province by BP neural network based on Bayesian regulation. The results show that the way of the neural network has a good predict effect and it is able to provide a reliable support for decision-making. |