A Study On The Price Evaluation Of Second Hand Houses Of Ordinary Residential Buildings In Chongqing Based On LightGBM Model

With the rapid development of the global economy,people’s demand for housing is also constantly changing,and second-hand housing has become the first choice for many people to invest and finance.In recent years,the real estate market has developed rapidly,occupying an important position in the national economy.People’s housing standards have not only been well improved,but also provided an important support for China’s growing economy.With the continuous development of the economy,the second-hand housing market has become increasingly active.Accurate estimation of second-hand housing prices can not only reduce transaction risks,protect the rights and interests of buyers and sellers,but also provide strong support for the healthy development of the real estate industry.However,currently,in domestic research on second-hand housing valuation,most of the techniques such as random forest and neural network are used,but due to technical limitations,the accuracy and efficiency of prediction still need to be improved.The purpose of this article is to use hedonic price theory and web crawler technology to construct a Light GBM model,taking second-hand residential housing in Chongqing as the research object,and estimate its value in a more accurate and efficient manner.The purpose of this article is to deeply explore the development and evolution of real estate valuation methods,clarify the purpose and meaning of scientific research,and summarize and analyze the achievements of domestic and foreign experts and scholars.Finally,from the four aspects of location,architecture,neighborhood,and time characteristics,we will select the factors that affect the price of ordinary residential second-hand housing.Using web crawler information technology,we will collect second-hand housing data from Chongqing Linked In for analysis,and through data cleaning and processing,Remove complex and unrelated data information and abnormal data information,and finally obtain complete second-hand housing information,in order to better understand the development trend of the real estate market and provide a reference basis for real estate valuation.Through in-depth analysis of feature sets using three different feature selection techniques,namely filtering,wrapping,and embedding,the optimal feature set is ultimately determined to achieve the best results.By using jupyter Notebook,we have established a more effective,objective,and applicable Light GBM price evaluation model for the general residential second-hand housing market in Chongqing.We divided the data set into two parts according to 4:1: a training set and a test set.The determination of the optimal parameters required by the model is trained using a divided training set,while the test set is used to verify the accuracy of the modeling.In order to verify the effectiveness of the model,the model is compared with traditional market methods,and two comparative valuation models,namely,random forest and XGBoost,are constructed for comparative analysis.Through cross validation,evaluation indicators such as determination coefficient,average absolute error,root mean square error,and average relative error percentage are used to comprehensively evaluate the effectiveness of the model.Through comparative analysis,it can be found that the model performance,prediction accuracy,and training speed of the Light GBM valuation model are optimal,and it is an excellent model that can be applied to the evaluation of second-hand housing prices for ordinary residential types in Chongqing.