论文部分内容阅读
Almon算法进行改造扩展可解决基于固定产业销售结构假定编制产业部门类型投入产出表出现负值的问题.根据Almon算法在迭代过程中避免负值的思想,基于固定产业销售结构假定下的推导公式,构造迭代公式,并针对迭代公式中可能出现负值的部分,引入比例因子,使在每一步的迭代中都能够避免负值,并且给出了迭代公式和比例因子的直观解释.为验证该方法的有效性,将扩展的Almon算法和传统负值处理方法进行实证比较,结果表明,扩展的Almon算法要明显优于传统负值处理方法,基于其编制出的投入产出表与官方公布投入产出表的误差更小.
Almon algorithm can solve the problem of negative value in input-output table of industrial sector based on the assumption of fixed-structure sales structure.According to Almon’s idea of avoiding negative values in iterative process, based on the derivation formula , Construct the iterative formula, and introduce the scale factor for the possible negative part in the iterative formula, so as to avoid the negative value in every iteration of the iteration, and give a direct explanation of the iterative formula and the scale factor. The effectiveness of the method is compared empirically with the extended Almon algorithm and the traditional negative value processing method. The results show that the extended Almon algorithm is obviously superior to the traditional negative value processing method. Based on the input-output table and the official announcement input, The error in the output table is smaller.