Machine-Precision Prediction of Low-Dimensional Chaotic Systems
Christof Schötz, Niklas Boers · 2025
Read Machine-Precision Prediction of Low-Dimensional Chaotic Systems by Christof Schötz, Niklas Boers (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-2507-09652.txt.
Description
Christof Schotz and Niklas Boers show that simple chaotic systems can be forecast to machine precision from clean data in a 2025 paper. The first author's name carries diacritics in the byline. Chaotic systems such as the Lorenz model are standard tests for methods that learn dynamics from data. Their method fits a regression on high degree polynomial features. It is computed in very high precision arithmetic and needs no knowledge of the system's equations.
For the Lorenz system, forecasts stayed valid far longer than in any prior work, and longer still with the best precision settings. The method matched the accuracy of standard solvers that do know the equations. Results held on a second, harder chaotic system with a non polynomial form, and extended to a higher dimensional spatiotemporal model. The authors suggest that forecasting low dimensional chaos from noise free data is effectively a solved problem.
Catalog record
| CREATOR | Christof Schötz, Niklas Boers |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 115 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-09652