Linear Regression Model of House Price in Boston
Volume 8, Issue 3, June 2020, Pages: 52-63
Received: Jun. 28, 2020;
Published: Jun. 29, 2020
Views 161 Downloads 34
Niu Yilin, School of Mathematics and Statistics, Lanzhou University, Lanzhou, China
Follow on us
The change of house price is a common phenomenon. People are eager to grasp the law of house price to become the winner of real estate investment. This paper uses Boston house price data to explore the relationship between Boston house price and which independent variables. This paper uses linear regression model to construct the relationship between housing prices and crime rate in Boston. First, the classical linear model is adopted. Then we do the collinearity test, removing the lever point and other operations, the residual of the model still does not conform to the normal distribution, so the classical linear model cannot describe the data very well. Then, we add the quadratic term and the cross term, and use the method of stepwise regression to get the optimal regression autoregressive quantum set. After removing the leverage point and significance test, we found that the residual distribution was approximately normal. It shows that the improved model has well described the law of data. Finally, according to the data, the main conclusions are as follows: house price and tax rate, index close to the highway and index close to the city center are inversely correlated, which is positively correlated with the number of rooms, the proportion of teachers and students, and whether it is close to the Charles River. In addition, the concentration of nitric oxide, the proportion of low-end population and crime rate also have a certain relationship with housing prices.
Changes in House Prices, Linear Regression, Stepwise Regression, Positive Correlation, Negative Correlation
To cite this article
Linear Regression Model of House Price in Boston, Science Discovery.
Vol. 8, No. 3,
2020, pp. 52-63.