Toward a physics-guided machine learning approach for predicting chaotic systems dynamics
Liu Feng, Yang Liu, Benyun Shi, Jiming Liu · 2025
Read Toward a physics-guided machine learning approach for predicting chaotic systems dynamics by Liu Feng, Yang Liu, Benyun Shi, Jiming Liu (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/pmc-11782262.txt.
Description
Liu Feng and co-authors propose teaching forecasting machines the physics of the systems they predict, in a paper whose text dates it to 2024. Chaotic systems, from epidemics to extreme weather, defeat forecasters because tiny starting errors explode over time. Pure data-driven methods fit the past well and shine briefly, then drift. They ignore the rules that actually drive the system, which caps their long-range accuracy.
The team's method, physics-guided learning, feeds those governing rules into the training itself. The model still learns from records, but its guesses are steered toward paths the true equations allow. The aim is to stretch the horizon over which a forecast stays trustworthy. The paper sets out the approach and its rationale for chaotic dynamics, a test bed where long-term prediction is famously hard-won.
Catalog record
| CREATOR | Liu Feng, Yang Liu, Benyun Shi, Jiming Liu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |