Fixed-order PCA: Theory for Overestimated Factor Models
Yuan Liao, Xin Tong, Wanjie Wang, Dacheng Xiu · 2026
Read Fixed-order PCA: Theory for Overestimated Factor Models by Yuan Liao, Xin Tong, Wanjie Wang, Dacheng Xiu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-18448.txt.
Description
Yuan Liao and three co-authors develop, in a 2026 paper, limit theory for principal component analysis of a high-dimensional factor model. The working size R is fixed and need only meet or beat r, the true factor count. From anisotropic local laws in random matrix theory, extra sample eigenparts past r prove noise-led, unlinked, and nearly orthogonal to factor loadings. Two rotations, wide and narrow maps, both give consistent factor estimates. A factor-boosted test case proves root-T limit normality for every fixed R.
Catalog record
| CREATOR | Yuan Liao, Xin Tong, Wanjie Wang, Dacheng Xiu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 137561 |
| LICENSE | http://creativecommons.org/publicdomain/zero/1.0/ |
| RIGHTS | CC0 |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.18448