基于BP神经网络的我国M2/GDP预测与对策

Prediction and Countermeasure of Chinese M2/GDP Based on Neural Network

本文计算了1993年至2005年我国M2/GDP,发现该数据总体呈上升趋势,这一定程度上反映了我国体制上的深层次问题。另外,本文运用带有动量项和自适应学习率的BP神经网络方法预测出我国2006年的M2/GDP是1.6386左右,并提出了一些相应的政策建议。

This paper found that the overall upward trend in the data through calculating Chinese M2/GDP（1993-2005）.This to some extent reflects the deep-rooted problems in our system. In addition, this paper predicts M2/GDP in 2006 is 1.6386 by using adaptive learning rate and momentum term with BP neural network, then brings forward some corresponding policy recommendations.