Extraction of Greatest Impact Factor in Nonlinear Diagnosis Models
暂无分享,去创建一个
For diagnosis models with one response variable influenced by multiple factors, the paper proposes a method that finds the greatest impact factor, referred to as main factor. The method is based on statistical analysis and uses the principal component transformation to optimize statistics. It includes several steps: sampling, calculating and constructing the correlation matrix between response variables and factors, obtaining the most relevant matrix by principal component transformation, determining main factor by comparing the correlation degree of correlation matrices and the most relevant matrix. The algorithm can meet the need of many engineering problems that hunt for the greatest impact factor in non-linear diagnosis models with multiple factors.
[1] Liang Bang. Application of multivariate statistical analysis approach to the appraisal of teaching quality , 2003 .
[2] Maria Petrou,et al. Image processing - the fundamentals , 1999 .
[3] Roberto Viviani,et al. Functional principal component analysis of fMRI data , 2005, Human brain mapping.
[4] Alois Kneip,et al. Common Functional Principal Components , 2006 .