基于NAR模型的上海市房产税规模预测  

Shanghai Property Tax Scale Prediction Based on NAR Model

作　　者：刘洋[1] 蔡明明[1] 杨婉莹[1] 

机构地区：[1]武汉大学资源与环境科学学院,湖北武汉430079

出　　处：《中国房地产》2016年第30期17-23,共7页China Real Estate

摘　　要：运用上海市2004-2015年的社会经济指标数据,对影响上海市房产税征收规模的11个指标进行主成分分析,将所得主成分综合得分引入NAR神经网络模型中来预测上海市2016-2020年房产税征收规模得分,并采用多元回归法对房产税规模与其得分间的关系进行拟合,得到上海市2016-2020年房产税征收规模预测值。研究结果表明,房产税规模与其得分的关系近似趋近于指数函数y=16.34e0.183x,上海市2016-2020年房产税征收规模增长呈逐年上升趋势,涨幅逐渐趋向平稳。此房产税规模预测方法可以运用到其他省市的房产税征收规模情况的计算中,为中国房地产市场宏观调控政策的实施提供依据。By analyzing Shanghai social economic indicators data from 2004 to 2015, this paper use PCA to research 11 indexes which affect Shanghai property tax scale. Put principal component comprehensive scores into the NAR neural network model to predict scores from 2016 to 2020, then use multiple regression to build relationship between the property tax levy and scale score to get the property tax levy predicted value from 2016 to 2020 in Shanghai. The results show that the relationship between the property tax levy and scale score approximates to exponential function and there is an upward trend in Shanghai property tax scale from 2016 to 2020 which gradually incline to stable. This property tax scale prediction method can be applied to scale calculation of other provinces and provide basis to the implementation of Chinese real estate market macroeconomic regulation and control policy

关 键 词：房产税 NAR神经网络 主成分分析 

分 类 号：F812.42[经济管理—财政学] F299.23