https://github.com/krassowski/loadings-similarity
Eigenvectors or loadings similarity approach for the selection of number of components in PCA
https://github.com/krassowski/loadings-similarity
eigenvectors pca principal-component-analysis singular-value-decomposition sklearn
Last synced: about 1 month ago
JSON representation
Eigenvectors or loadings similarity approach for the selection of number of components in PCA
- Host: GitHub
- URL: https://github.com/krassowski/loadings-similarity
- Owner: krassowski
- License: mit
- Created: 2020-03-22T22:37:47.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2020-03-23T15:06:33.000Z (about 6 years ago)
- Last Synced: 2025-01-13T18:46:07.491Z (over 1 year ago)
- Topics: eigenvectors, pca, principal-component-analysis, singular-value-decomposition, sklearn
- Language: Jupyter Notebook
- Size: 891 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# loadings-similarity
Eigenvectors or loadings similarity approach for the selection of number of components in PCA
Accompaning article: [Notes on the number of components in PCA: R², Q² & eigenvectors similarity](https://towardsdatascience.com/notes-on-the-number-of-components-in-pca-r%C2%B2-q%C2%B2-eigenvectors-similarity-60a19b2f671a?source=friends_link&sk=ae54130d659ffb448aee433ea98994c3).
Please leave a comment if you find this method helpful (if it is of general interest, I will publish further validation benchmarks and a ready-to-use Python package).