The Application Of Partial Linear Model In The Rental Housing Price Prediction

With the rapid development of the Chinese economy,the process of the housing commercialization and the urban modernization has dramatically sped up.As the main body of the real estate market,the housing rental market could cool down the continuously overheated real estate prices,and it also relieves the pressure for the people who have housing needs but without purchasing ability.However,the housing market,now,arrests the most attention,while the research aiming at the housing rental market,especially the rental price forecast,is very lacking.In facing the complex and diversified housing rental market and residents' gradually diversified and personalized living needs,this paper adopts scientific methods to forecast the housing rental price.It also helps the tenants to find the living space that meets their own needs efficiently and quickly,and provides direction guidance for urban reconstruction.The main work of this paper is as follows:First,the paper introduces the current situation of the housing rental market,on this basis,summarizes the existing research on the factors affecting the housing rental market rent,and analyzes the innovation points and shortcomings of the current research methods.Second,this paper summarizes the Partial Linear Model and elaborates on the application of the Partial Linear Model.Then it also accordingly proposes the estimation method and test method of the Partial Linear Model and introduces theoretical knowledge of Principal Component Analysis.Third,it summarizes the seven categories that affect the housing rental price,including 42 indicators.A total of 15 main influencing factors has selected after the preliminary research.Through the urban real estate big data platform and Fangtianxia rental website,this paper obtained 693 rental information data and quantified the research data.Fourth,the method of Principal Component Analysis has applied in Dimensionality Reduction for processing the screened indicators data.And four principal components are shown after this process.Then,through fitting of the Partial Linear Model and Linear Regression Model,this article forecasts the housing rents on the website and compares the predicted results of the two methods with the actual housing rent to indicate that the Partial Linear Model has application value.