改善的ADL模型对中国GDP增长率的研究与预测

Research and Forecasting on Chinese GDP Growth Rate Based on Improved ADL Models

本文提出三种创新性模型PLSADL,RADL和RPLSADL.这三种新模型是将考虑了数据时序性的时间序列ADL模型与考虑了多变量共线性问题的多元线性回归模型PLSR,RR,RPLS相结合.通过分析我国2000年到2012年季度GDP增长率与八项经济指标的关系,我们发现新模型PLSADL,RADL和RPLSADL在拟合效果和预测能力上都优于其它四个模型.这说明在ADL模型的建立过程中,如果能够考虑多变量共线性问题将会有效地提高模型的预测效果.

We suggest three new models including PLSADL, RADL and RPLSADL. These models are ensemble approach to GDP time series forecasting which integrating au- toregressive distributed lag （ADL） model with partial least squares regression （PLSR） ,ridge regression （RR） and ridge partial least squares regression （RPLSR）. The ADL model consid- er the property of the time series data and PLSR,RR and RPLSR are applied to reduce the number of variables used in ADL model. The performances of proposed new models show the satisfied results not only in fitting on training data but also in predicting on test data.