提高工业产值灰色模型预测精度

Improve the Forcasting Accuracy of the Grey Model of Industrial Output Value

通过建立鞍山市历年工业总产值全数据的灰色模型,较准确地预测出鞍山市1990年的工业总产值,其预测误差为10.99％.在1989年总产值的邻域系上建立了灰色模型簇。通过比较簇中各灰色模型的平均误差及原点误差,筛选出最优灰色模型。为进一步提高该模型的精度,又以残差数列建立灰色模型对该模型进行修正。修正后预测出1990年的工业总产值的预测误差仅为4.14％,这表明,在灰色模型簇中筛选最优预测模型,并在此基础上建立被残差修正后的灰色预测模型,从而提高灰色模型的预测精度。

From the grey model of the gross value of industrial output of Anshan over the years, We can forecast the gross value of industrial output of 1990, the forecasting error is 10. 99%. In the nei-bourhood of the gross value of 1989. We build the family of the grey models. From compare theaverage error and origin error of every model of the family, we select the optimal model. In orderto improve the accuracy of the model, we modify it again suing a grey model established by seriesof residual. From the modified forecasting model. We forecast the gross value of industrial outputof 1990. It's forecasting error is only 4.14％. This illustrates that from the family of grey models selected the optimal forecasting model and modified this model. We can improve the forecasting ac-curacy of the grey model.