Application Research Of PCA-SVR Model For Real Estate Tax Base Evaluation Model

Shortly after the end of the 19 th National Congress of 2017,Xiao Jie,the then Minister of Finance,wrote in the People 's Daily that he clearly pointed out the need to promote property tax legislation and implementation in accordance with the principle of“legislation first,full authorization,and step-by-step advancement”.This is another official announcement of the reform of the property tax system since the real estate tax reform in Chongqing and Shanghai in 2011.The comprehensive collection of real estate tax is conducive to the improvement of China's existing property tax system to better build a socialist tax system with Chinese characteristics.Secondly,the comprehensive collection of real estate tax is also of certain significance to suppress the current overheating of the real estate market and excessive investment.In recent years,the state 's control over the real estate market has become stricter.The Central Economic Work Conference at the end of 2018 clearly stated that“the house is used for living,not for speculation”.The July 30,2019 Politburo meeting proposed “ No The assertion that real estate is used as a short-term stimulus to the economy once again reflects the central government's determination to control the real estate market.To fully levy real estate taxes,we must first determine the real estate tax base assessment problem.At present,China does not have specific tax base assessment criteria and specific methods.Most of them are still in the theoretical exploration stage,so we should start from China to construct a tax that meets national conditions and meets local characteristics The basic evaluation model has stronger practical significance.Based on theoretical analysis and empirical analysis,this paper starts from the current research status of tax base assessments at home and abroad,and clarifies that the tax base of the current real estate tax base assessment is mainly the value of the real estate ownership stage,and then clarifies the tax base value from related theories at home and abroad standard.Starting from the current status of China's tax base assessment,combined with foreign practical experience and the results of domestic research and exploration,it is clear that batch assessment is the most effective method for tax base assessment.Then,Chongqing actually set up a feature variable that combines the main factors affecting the value of the second-hand housing market.The index system,theselected index is brought into the PCA-SVR model for training and prediction,and a good prediction effect is obtained,Better fitting effect than neural network.Therefore,it provides new ideas and research directions for the comprehensive tax base assessment of real estate taxes in the future.