An Empirical Bernstein Inequality for Dependent Data in Hilbert Spaces and Applications
Erfan Mirzaei, Andreas Maurer, Vladimir R. Kostic, Massimiliano Pontil · 2025
Read An Empirical Bernstein Inequality for Dependent Data in Hilbert Spaces and Applications by Erfan Mirzaei, Andreas Maurer, Vladimir R. Kostic, Massimiliano Pontil (2025) 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-2507-07826.txt.
Description
Erfan Mirzaei and three co-authors prove an empirical Bernstein inequality for dependent data in Hilbert spaces, in work from 2025. Learning often must cope with data that are neither independent nor identically spread. Their bounds are data-dependent and fit vector-valued processes, for both stationary and non-stationary cases, using fast decay of links between far-apart times.
Two uses show the value of the bounds. One is covariance operator estimation in the Hilbert-Schmidt norm, and the other is operator learning for dynamical systems, each with new risk bounds. Numerical tests in both settings show what the bounds mean in practice. The work widens classic concentration tools to a much broader data world.
Catalog record
| CREATOR | Erfan Mirzaei, Andreas Maurer, Vladimir R. Kostic, Massimiliano Pontil |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 89 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-07826