The Research of Multiple Regression Analysis in Rural-Urban Income Disparity

Jian Li, Xiangyu Guo
2016 International Journal of Smart Home  
The multiple linear regression model contains more than one predictor variable and it shows the relationship among multiple variables. In the existing research field of ruralurban income disparity, the method of multiple regression analysis is mainly employed. But the linear relationship among variables is estimated mainly depending on principal component analysis. Principal component analysis is used to convert a set of observations of possibly correlated variables into a set of values of
more » ... t of values of linearly uncorrelated variables called principal components. The principal component analysis is widely used for feature extraction to reveal the most main factors from the multiple aspects. A multiply linear regression model integrating principal components analysis is proposed to address on the income gap between the city and country. The influential factors are given and the analysis results are discussed in this paper. The experimental results on income data from 1990 to 2013 show that the proposed method is effective in predicting the income ratio and analyzing the influential factors. urban income disparity, and (2) reveal the correlated influential factors of rural-urban income disparity. This paper chooses nine features including urbanization level, rate of supporting agriculture in finance, growth rate of GDP et al. By using the principal component analysis, we construct the regression model to predict the income ratio. And the correlated influential factors are also analyzed by using the proposed method. This paper is organized as followed. In Section 2, we discuss the multiple linear regression model and the principal component analysis is introduced. In Section 3, the regression model integrating principal component analysis is constructed. In Section 4, we give an example analysis by using rural-urban income data from 1990 to 2013. And our method is concluded in Section 5.
doi:10.14257/ijsh.2016.10.11.19 fatcat:qhs4xl26xrcjle4hiiyvf3yfey