Prediction Research of the Per Capita GDP of Liaoning Province Based on the Improved BP Algorithm

Because the time series of the per capita GDP is of complex and nonlinear characteristics,traditional forecasting methods tend to produce larger errors.The paper establishes and trains the neural network prediction models by different improved algorithms based on the Matlab software to observe the accuracy and effectiveness of algorithms.The paper finally predicts the per capita GDP of Liaoning province by using this model.