Research On Chongqing Second-hand House Valuation Model Based On Random Forest Model

As one of the pillar industries of China's national economy,the real estate industry,under the background of rising house prices,is under pressure for ordinary buyers with rigid demand.House prices have become a hot topic for a long time,and have received great attention from the government,society and academia.Due to the continuous acceleration of the urbanization process in China,the land available for development in cities has gradually decreased,and the second-hand housing market has entered people's attention.Due to the increased demand for second-hand housing information,it is of great significance to study the price of second-hand housing.Due to the complexity and non-linear relationship of influencing factors on real estate prices,it is difficult for traditional assessment methods to accurately assess the value of real estate.How to establish more effective indicators and models to ensure the accuracy of the evaluation and to estimate the price of second-hand housing efficiently and accurately has become an urgent need for research topic.Based on the above background,this article takes 270 ordinary second-hand housing houses in the 9th district of Chongqing as the research object,and focuses on completing the following tasks:(1)Based on the work of previous scholars,through reading a large number of literature,understanding of the real estate evaluation theory forming a systematic framework,and analyzing the current situation of the use of machine learning in the evaluation market,clarifying the research background,significance,and content of the thesis method.(2)Explain the working principle of Random Forest and compare it with other widely used machine learning algorithms.Finally,the advantages of random forest algorithm are obtained.(3)Select second-hand housing in the 9th district of the main city of Chongqing as the research object,find relevant information about the house,summarize and organize effective information,and normalize the data.Based on the previous scholars' research,combined with the sample of the paper,the maximum Find and screen the effective housing evaluation indicators for this article.The index models are divided into three categories: architectural characteristics,neighborhood environment,and location characteristics,including light rail,public transportation,distance to the CBD,school district housing,housing area,age,floor,There are 17 indicators of the number of bedrooms,orientation,decoration,floor area ratio,greening rate,quality of property management,neighbor interference,environmental sanitation management,commercial facilities and infrastructure facilities.(4)Debug out the two key parameter values of the model construction,and get the two optimal parameters of Chongqing second-hand housing price evaluation model Mtry value of7,and Ntree value of 150.Finally,through a comparative analysis of the evaluation results of the second-hand house price of the multiple linear regression model and the evaluation results of the random forest model,it was found that the random forest algorithm has higher prediction accuracy and performs better in the empirical research of property price prediction,Better forecasting performance is more suitable for real estate price evaluation,and provides a new technology that can be used and promoted for modern real estate evaluation.