Research On The Price Evaluation Model Of Second-hand Housing In Jinan Based On Machine Learning

In 2021,the transaction volume of commercial housing across the country showed a trend of high first and then low.In many places,the "reference price of second-hand housing" is mentioned for the first time to guide price expectations and curb investment demand.In the second-hand house transaction process,because consumers and sellers have different demand for house prices,it is difficult to reach an agreement on the price of second-hand houses.Therefore,it is necessary to effectively evaluate the price of second-hand houses.It can bring more accurate price information to homebuyers,intermediary companies and second-hand home suppliers,thereby improving the accuracy of second-hand home price assessment.There are many methods for valuing second-hand houses.The traditional price evaluation methods focus on qualitative analysis and are greatly influenced by subjective factors.With the progress of society,many scholars have begin to use machine learning methods to evaluate real estate prices.This method focuses on quantitative analysis and is more accurate.Referring to the housing price evaluation methods in recent years,this paper uses the machine learning method to evaluate the second-hand housing prices in Jinan,and achieves good results in the empirical analysis.This paper selects 17 characteristic variables such as building area and decoration situation to explore their correlation with second-hand housing prices.First,descriptive statistical analysis is carried out on the influencing factors of second-hand housing prices in Jinan.Then corresponding explanation,screen and quantify the various characteristic variables that affect the price of second-hand housing,draw on the theory of hedonic price,divide the architectural,location and environment factors of second-hand houses,and build MLR,SVM,RF and XGBoost,and then adopt the Stacking strategy to use random forest,support vector machine and XGBoost as the basic model,and use multiple linear regression as the final evaluation model to evaluate the price of second-hand houses.Finally,the evaluation indicators such as mean square error and absolute value error are used to evaluate the feasibility of models.The research shows that the second-hand housing area is closer to the subway or bus station,the higher the price;the house faces south and is on the middle floor and the price of hardcover houses is relatively high;in addition,the use of fusion models to evaluate the prices of second-hand houses has high accuracy and is worthy of promotion and application.