Means and Issues for Adjusting Principal Component Analysis Results
Algorithms · 25.02.2025 · 18 (3) · 129
DOI: 10.3390/a18030129 ↗Kurz gefasst
Fachartikel in der Zeitschrift Algorithms, online erschienen am 25.02.2025. Verfasst von Tomokazu Konishi. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
Principal component analysis (PCA) is a method that identifies common directions within multivariate data and presents the data in as few dimensions as possible. One of the advantages of PCA is its objectivity, as the same results can be obtained regardless of who performs the analysis. However, PCA is not a robust method and is sensitive to noise. Consequently, the directions identified by PCA may deviate slightly. If we can teach PCA to account for this deviation and correct it, the results should become more comprehensible.
Methods
The top two PCA results were rotated using a rotation unitary matrix.
Results
These contributions were determined and compared with the original. At smaller rotations, the change in contribution was also small and the effect on independence was not severe. The rotation made the data considerably more comprehensible.
Conclusions
The methods for achieving this and an issue with this are presented. However, care should be taken not to detract from the superior objectivity of PCA.
Zitieren
APA
Konishi, T. (2025). Means and Issues for Adjusting Principal Component Analysis Results. Algorithms, 18(3), 129. https://doi.org/10.3390/a18030129
Vancouver
Konishi T. Means and Issues for Adjusting Principal Component Analysis Results. Algorithms. 2025;18(3):129. doi:10.3390/a18030129
BibTeX
@article{Konishi2025,
author = {Konishi, Tomokazu},
title = {Means and Issues for Adjusting Principal Component Analysis Results},
journal = {Algorithms},
year = {2025},
volume = {18},
number = {3},
pages = {129},
doi = {10.3390/a18030129}
}
EUVIMED