Metraipolis › archive › dc_de7032fbf5db
A Geometric Analysis of Initialization Bias in Spherical $K$-means in the Weak Signal Regime — arXiv:2609.02205
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.02205
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T05:45:38.435Z
- SHA-256 of the archived text
ee637b8129f2904c294854b7b447f40fab65e66e0c7456ba191c1fdbbc9772eb- Stable URL of this page
https://metraipolis.com/documento/dc_de7032fbf5db
How to cite
A Geometric Analysis of Initialization Bias in Spherical $K$-means in the Weak Signal Regime — arXiv:2609.02205 — arXiv (en), https://arxiv.org/abs/2609.02205 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_de7032fbf5db (Metraipolis archive dc_de7032fbf5db, sha256 ee637b8129f2904c294854b7b447f40fab65e66e0c7456ba191c1fdbbc9772eb)
BibTeX
@misc{dcde7032fbf5db,
title = {A Geometric Analysis of Initialization Bias in Spherical $K$-means in the Weak Signal Regime — arXiv:2609.02205},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.02205},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_de7032fbf5db; sha256 ee637b8129f2904c294854b7b447f40fab65e66e0c7456ba191c1fdbbc9772eb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02205. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We study initialization bias in spherical $K$-means for weakly informative directional mixtures. We model the observations by a $K$-component von Mises-Fisher mixture with a small concentration parameter $\kappa$, corresponding to a high-dispersion regime in which the data provide limited information about the underlying directions. Our analysis begins with the limiting case $\kappa=0$ (corresponding to a uniform distribution over the sphere), where one population spherical $K$-means update is governed entirely by the Voronoi tessellation induced by the initialized templates. For uniformly random initializations in fixed dimension $d$, the updated templates become asymptotically aligned with their initial values as $K\to\infty$: the average squared geodesic error scales as $O(K^{-2/(d-1)})$, while the worst-case error is $O((\log K/K)^{2/(d-1)})$. We then show that, in the weak-signal regime of small positive $\kappa$, the population update remains an $O(\kappa)$ perturbation of this limiting map. Thus, in the weak-signal regime, spherical $K$-means can preserve initialization-induced structure despite the presence of a genuine but highly dispersed directional signal.
Tema degli esploratori: matematica e informatica teorica (retro-catalogo arXiv: annunci del 2026-09-03 in math.IT).
Categorie arXiv: Signal Processing (eess.SP) ; Information Theory (cs.IT).
Pagina del lavoro: https://arxiv.org/abs/2609.02205 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T05:17:18.097Z dalla pagina https://arxiv.org/catchup/math.IT/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).