论文部分内容阅读
由于区域经济系统中许多经济变量呈现出强非线性与大波动性的特征,使得传统的时间序列线性建模和预测技术难以适应区域经济预测的要求.为此,提出基于支持向量机改进的残差自回归区域经济预测模型.首先采用时间序列分析中的残差自回归模型对时间序列趋势进行线性拟合,然后对残差自回归模型估计后的残差序列采用支持向量回归方法再次提取其非线性特征,从而提高区域经济时间序列模型的预测精度.最后以广东省GDP的预测实例说明模型的有效性.
Because many economic variables in the regional economic system are characterized by strong nonlinearity and volatility, the traditional time series modeling and forecasting technology are difficult to meet the requirements of regional economic forecasting.Therefore, the improved residuals based on support vector machines Regression autoregressive regional economic forecasting model.Firstly, using the residual autoregressive model in time series analysis to linearly fit the time series trend, the residual autoregressive model was estimated by using the support vector regression method to extract the residual sequence again Non-linear characteristic, so as to improve the prediction accuracy of regional economic time series model.Finally, the effectiveness of the model is illustrated with the example of GDP forecast in Guangdong Province.