基于空间自回归性的低保标准模型的建立

Establishment of Multiple Minimum Living Standard Model Based on Spatial Autoregression

为实现城乡低保标准的调整工作更加规范,需要参照扶贫标准和全国平均低保标准进行论证和科学的测算,现有低保标准对居民经济社会发展水平和财政承受能力了解的还不够全面,因此建立一种“基于空间自回归性低保标准”的数学模型来解决以上问题.通过统计分析方法对其进行推理假设,确定最优计算指标,进行针对性查找,将主要指标划分为8个影响因素,发现相邻两个地区的低保应由本地区的经济发展状况和相邻地区对本地区的经济影响两个因素决定.通过回归评价指标RMSE及MAE对模型性能进行评价,归一化处理后的空间自回归模型RMSE为0.0477MAE为0.1625,在两个指标上同时具有较好的预测性能及鲁棒性.

In order to make the adjustment of urban and rural subsistence allowances more standardized,it is necessary to demonstrate and scientifically calculate with reference to the poverty alleviation standard and the national average subsistence allowance standard.The existing subsistence allowance standard does not fully understand the residents’economic and social development level and financial affordability.Therefore,a mathematical model based on spatial autoregressive subsistence allowance standard is established to solve the above problems.Through the statistical analysis method,the inference hypothesis is made,the optimal calculation index is determined,targeted search is carried out,and the main indexes are divided into 8 influencing factors.It is found that the minimum living standard in the two adjacent regions should be determined by the economic development status of the region and the economic impact of the adjacent regions on the region.The performance of the model is evaluated by regression evaluation indexes RMSE and MAE.The normalized spatial autoregressive model RMSE is 0.0477 and MAE is 0.1625.It has good prediction performance and robustness in both indexes.