纠正传统回归误差的半参数变系数估计模型设定及应用

Setup and Application of Semi-parametric Variable Coefficient Estimation Model of Correcting Traditional Regression Error

针对传统回归模型对事物空间相关性估计可能出现的误差,文章构建了半参数变系数空间误差模型分析区域金融资源空间配置和经济发展之间的相关性,基于大样本对所构建模型的渐近正态性和一致性进行检验,变化窗宽使用蒙特卡洛方法对模型的有效性进行模拟,实证研究区域金融资源空间配置与经济发展的相关性。

Aiming at the possible errors in using the traditional regression model to estimate the spatial correlation, this pa- per constructs a semi-parametric variable coefficient spatial error model to analyze the correlation between spatial allocation of re- gional financial resources and economic development. And then, based on the large sample, the paper makes a test on asymptotic normality and consistency of the established model. Finally the paper use Monte Carlo method to simulate the effectiveness of the model by changing the window width, and empirically studies the correlation between regional spatial allocation of financial re- somces and economic development.