基于矩阵转换技术的区域投入产出表预测——以浙江为例

Forecasting the Regional Input-Output Tables Based on Matrix Transformation Technology——Taking Zhejiang Province as the Sample

官方投入产出表的间断性和时滞性限制了投入产出模型在产业关联动态分析中的应用。本文基于矩阵转换技术、成分数据分析方法与时间序列预测模型,设计了区域投入产出表三个象限数据的预测方法及预测效果评估方法,并以浙江为例预测了2005—2017年39个部门的投入产出序列表。本文研究显示:(1)设计的投入产出序列表预测方法简易可行,且充分运用了宏观经济统计数据;(2)详细呈现出了投入产出序列表的预测过程,为区域投入产出序列表的实际预测提供了借鉴;(3)预测得到了浙江投入产出序列表,可为产业关联及部门影响动态分析提供基础数据。

The discontinuity and time lag of official input-output table have limited the application of input-output model in industrial correlation dynamic analysis.Based on matrix transformation technology,component data analysis method and time series prediction model,this paper designed the prediction method and prediction effect evaluation method of the three quadrant data of regional input-output tables,and took Zhejiang province as an example to predict the input-output sequence tables of 39 sectors from 2005 to 2017.The research covered as follows.Firstly,the input-output sequence tables prediction method designed in this paper was simple and feasible,and made full use of macroeconomic statistical data.Secondly,the prediction process has presented the input-output sequence tables in detail,which has provided references for the actual prediction of regional input-output sequence tables.Finally,the input-output sequence table of Zhejiang has been predicted,which may provide basic data for the dynamic analysis of industrial correlation and departmental impact.