Bounded overestimation of the number of factors in PCA preserves √T-valid inference and consistent factor-space recovery under a random-matrix local law.
arXiv preprint arXiv:2406.13635 , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
LAE-EnKF learns a nonlinear encoder-decoder pair plus a stable linear latent dynamics operator so that the Kalman filter can be applied entirely in latent coordinates.
citing papers explorer
-
Fixed-order PCA: Theory for Overestimated Factor Models
Bounded overestimation of the number of factors in PCA preserves √T-valid inference and consistent factor-space recovery under a random-matrix local law.
-
Latent Autoencoder Ensemble Kalman Filter for Nonlinear Data assimilation
LAE-EnKF learns a nonlinear encoder-decoder pair plus a stable linear latent dynamics operator so that the Kalman filter can be applied entirely in latent coordinates.