Study on Linear Regression Analysis of Multivariate Compositional Data
Compositional data analysis is used to reflect the proportion structure of research objects. In this paper constrained regression is applied to compositional data regression and a method of multivariate regression analysis, based on the inner product in space Rnxm, is put forward, which aims at the situation when dependent and independent variables are all correlated compositional data. This method transforms the unit-sum constraint within components into constrained regression, and eliminates the adverse effects from multicollinearity. At the same time, by keeping all the information from original variables, the method ensures the regression model can be explained by compositional data with different meanings.
compositional data inner product based on space Rnzm multivariate compositional data linear regression linear constrained regression
Guo Lijuan Yi Bin Guan Rong
School of Economics & Management, Beijing University of Aeronautics and Astronautics, Beijing, China
国际会议
The 6th International Conference on Partial Least Squares and Related Methods(第六届偏最小二乘及相关方法国际会议)
北京
英文
389-393
2009-09-04（万方平台首次上网日期，不代表论文的发表时间）