Research On Application Of Economic System Based On PLS Regression Method

Partial Least-Squares is a new multivariate analysis method to build the equation about regression,which was fist proposed by economist S.Wold in 1983.Partial Least-Squares Regression was increasingly concerned later.Many statisticians study this method.The method is not only used for sigle variable regression anlysis,but also multiple analysis.Partial Least-Squares Regression is equivalent to the combination of multiple linear regression analysis, principal components analysis ,and canonical correlation analysis.(PLSR≈multiple linear regression analysis+princinpal component analysis +canonical correlation analysis). When multicollinearity is exist or the number of sample capacity are less than the number of independent variables,the new explanatory variables are built.The new variables use OLS to determine the equation between the dependent variables and independent variables.The construction of new explanatory variables of PLS take the most information and eliminate the multicollinear effects,it is good to predict the dependent variables.At the same time,the componets which the numbers are less the numbers of independent variables are used to reduce the dimensions about the problem of regression.This paper summarizes the multivariate linear regression and realizes the problem about multicollinearity,proposes the definition of Partial Least-Squares Regression (PLSR) model.The PLSR model is proposed by the combination of multiple linear regression model ,the principal components analysis and the canonical correlation analysis.The theoretically analysis indicates that they can offer more scientific and more reasonable than the traditional methods. Applying partial least-squares (PLS) regression in the final chapter of the paper.This paper analyzes the impact of 4 factors on the househould consumption of Heilongjiang Province of China.The result shows that the capability of househould consumption of Heilongjiang Province has relations with the price.,employment, the income of country,and the income of city. We have to resolve these questions to improve the household consumption.In addition, the PLSR model is applied to the compositional data after symmetric logratio transformation. A method of linear regression model is proposed when the dependent variables and independent variables are all correlated compositional data to establish model.For compositional data, this modeling method can satisfy the unit-sum constraint and eliminate the harm that derives from the complete multi-collineation.It emphasizes the function and significance of compositional data in modeling. On the basis of the new method built the employment structure model.