R-PCA replaces Euclidean distances in PCA with UMAP-induced local weightings and then performs PCA on a rescaled covariance; it claims better cluster preservation on a simulated set and Olivetti Faces.
Lapan, Irene E
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Riemannian Principal Component Analysis
R-PCA replaces Euclidean distances in PCA with UMAP-induced local weightings and then performs PCA on a rescaled covariance; it claims better cluster preservation on a simulated set and Olivetti Faces.