The Study Of Employment Problem In Our Country Based On Sparse Principal Component And Principal Component Regression

Currently,the employment problem has become an important issue.This paper first introduces the influential factors for employment,then selected eleven economic variables from China Statistical Yearbook,such as gross domestic product,resident consumption level,fiscal revenue,financial expenditure,tax revenue,fixed-asset investment,net exports,money supply,total energy consumption,R&D,expenditures,gross payroll,which are independent variables.The employment is the dependent variable,first,using the stepwise regression to establish model.However,when analyzing we find that model exist serious collinearity.Then we use the principal component regression(PCR)to analyze data.Principal component is the linear combination of all the economic variables,which is difficult to explain the importance of each economic variable.While the sparse principal component analysis(SPCA)can more clearly show the impact of economic variables on the number of employment.However,the cumulative contribution rate of sparse principal component is lower than that of general principal component,and the accuracy of prediction is lower than that of the PCR model when establishing the regression model of employment.Therefore,PCR is more reasonable and accurate in the employment forecast.In the analysis of important factors affecting the employment,more satisfactory results can be obtained through SPCA.