A note on principal component analysis f
β
Jianguo Sun
π
Article
π
2000
π
Elsevier Science
π
English
β 77 KB
We consider the application of principal component analysis (PCA) to the analysis of high-dimension data with the analysis goal being calibration. Two commonly used versions of PCA are compared and it is showed that contrast to the expected, the simpliΓΏed version could underestimate prediction error