基于BP神经网络的西安市宏观经济预测

宏观经济系统是一个复杂的非线性系统,对宏观经济进行预测应采用非线性的工具进行建模。采用BP神经网络对西安市宏观经济指标进行预测,此预测模型只需少量训练样本就可以确定网络的权值和阈值。实验表明模型预测精度高,能够对西安市宏观经济系统中的非线性关系进行描述,使建立的非线性模型与实际系统更加接近。

The Xi'an City Macroeconomic Prediction Based on BP Neural Network

Macroeconomic system is a complex nonlinear system .The non-linear tools should be used in macroeconomic models. In this paper, BP neural network is used in forecasting the macroeconomic indicators. The forecasting model needs a small number of training samples to determine the network weights and thresholds. Experiments show that there exists high prediction accuracy. It can describe the non-linear relationship existing on the macroeeonomie system .Through doing this, the non-linear model is closer to the actual system.