A PLS Method Based on Schmidt Transformation of Multi- block Independent Variables: an Application

摘要

We have proposed a PLS method based on a Schmidt transformation in order to select variables from multi-block independent variables. At first, the Schmidt method is applied to select the major independent variables in each block; then a PLS regression is implemented on the selected variables to overcome the collinearity problem and which guarantee the fit and explanatory power of the model. In the end, the method was applied to capture factors affecting timber consumption as a case study to illustrate the efficiency and rational of the method.