Research On The Price Index Of Residential Commercial Housing In Guangzho

In recent years,it has been nearly three years since the epidemic swept the world.It has had a far-reaching impact on the Chinese economy in many aspects,intensified downward pressure on the Chinese economy and disrupted the existing pace of macro-control.What is worth mentioning is that some objective problems that China has been facing for a long time under the pandemic have become more evident,among which the downward trend of the real estate industry has further confirmed that real estate will no longer be the main engine driving economic growth.Therefore,it is of great significance to study the real estate price index in the new environment.This paper mainly studies the housing price index with the help of relevant transaction data during the epidemic period of residential commercial housing in Guangzhou.The contents are as follows :(1)the theory of hedonic price model at home and abroad,model construction and real estate price index preparation methods are systematically sorted out.(2)Conduct descriptive statistical analysis on the data to preliminarily understand the influence of characteristic variables on the transaction price of residential commercial housing.(3)The four traditional Hedonic models were used to conduct empirical research on the data.Stepwise regression method and principal component analysis method were used to screen the model variables,and then the influence of characteristic variables was analyzed.(4)The quarterly price index is prepared based on the traditional Hedonic model and random forest model,and the gray prediction and machine learning alignment index are used to forecast the quarterly price index.The empirical results show that :(1)the residuals of linear regression and semilog model in the traditional four Hedonic models are heteroscedastic,which is not suitable for studying the influence of characteristic variables on transaction price.(2)The double log model and Box-Cox model can well explain the influence of each characteristic variable on the transaction price after transformation.(3)In the process of principal component analysis,it is found that the higher the transaction unit price of each region in Guangzhou,the greater is the comprehensive degree of influence of characteristic variables,and the smaller is the vice versa.(4)The hedonic price model adopted and the hedonic price and its index studied by the random forest method are basically consistent with the laws of the real estate market.According to the results of the study,the following points are proposed :(1)positioning products reasonably according to their own needs.(2)Straighten their own mentality,clear purpose of house purchase.(3)Pay attention to national policies and understand the market situation.