Study on exploratory factor analysis in measurement information extraction and reconstruction
Dengchao Feng, Nan Xiang, Sun Wei, Xuhong Ren, Zhenyu Jing, Qing Guo · 2011
In order to explore the potential association of various factors in actual measurement data, the principle of exploratory factor analysis was discussed in the paper and its application for the measurement information mining was explored to implement the synthesis and simplification of original multivariate plane data. Firstly, the correlation analysis of original variable and the extraction of common factors were executed. Secondly, the factor matrix was calculated to obtain the physical significance of common factors. Finally, the factor score coefficient matrix was obtained by virtue of the regression analysis, which can combine and synthesize the relevant variable to realize the extraction and reconstruction of measurement information. The simulation experiment testified the validity of the model. However, the reasonable selection of the extraction method of common factor and the quantitative criterion still need further research.