Statistical Analysis Of Income And Consumption In China

With the development of economy of our country, the income and consumption of residents have come to be a central issue gradually. So there is an increasing number of researchers study the relation between income and consumption with variety kinds of method. As a result, a series of theories and models have been put forward, which provide theoretical guidance and research methods for the study in turn. In this paper, we analyze the income and consumption of residents by using statistic theory and method. Also, we sign out some problems of the income and consumption and give some reasonable solutions for these problems. Firstly, listing out the change of income and consumption explicitly by analyzing income and consumption of some provinces and cities in our country with descriptive statistics method. Secondly, with the application of clustering analysis, classifying the provinces and cities into different groups and displaying all the categories they sorted to. At last, we apply the canonical correlation analysis to calculate the correlation of income and consumption. Again, we use the method of regression analysis to fit our country’ residential income and consumption, and according to the result of regression testing, we test the model to see whether it conforms with the assumptions of multivariate linear regression model, This article mainly aims at dealing with the multicollinearity, we apply the biased estimation of the principal component regression, ridge regression and stepwise regression, LASSO, elastic net method to model selection. After comparing the advantages and shortcomings of each model, we select out the best model, and give the economic interpretation of he parameters. Finally, based on the results of the analysis of people’s income and consumption, we make some rationality suggestions.