【摘要】 经济增长是一个多变量因素影响,具有复杂的非解析函数关系的系统。本文将人工神经网络用于经济增长的建模与预测之中,采用滚动优化技术把样本数据按时间序列化,使网络在训练过程中不断学习新的信息,提高网络的预测精度。经对河南省经济增长的模拟与预测,验证了该方法的可行性。更多还原

【Abstract】 The economic growth is a complex non-analytic function system influenced by many factors.This paper tries to use artificial neural networks in the modelling and the forecast of economy growth,and use the techniques of rolling optimization to arrange the sample data according to the time sequence so that it is possible to let the network study the recent information unceasingly in the training process and to enhance the forecast precision of the network.Having simulated and forecasted the economic growth of Henan province,it is turned out to be feasible.更多还原