On the Necessity of Two-Stage Estimation for Learning Dynamical Systems under Both Noise and Node-Wise Attacks
Jihun Kim, Javad Lavaei · 2026
Read On the Necessity of Two-Stage Estimation for Learning Dynamical Systems under Both Noise and Node-Wise Attacks by Jihun Kim, Javad Lavaei (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-2602-07288.txt.
Description
Identifying a dynamical system from one trajectory is hard when the data carries usual noise, adversarial attacks, or both at once. Jihun Kim and one co-author develop robust identification frameworks for linearly parametrized non-linear systems that handle each regime.
For the agnostic case, where the regime is unknown, the Huber estimator achieves a fast error rate under noise and a bounded error under attack. It acts as a bridge between mean- and median-based methods. For concurrent noise and attacks, the paper shows no convex one-stage estimator suffices and builds a two-stage option. Control theorists get robustness guarantees matched to the corruption they face.
Catalog record
| CREATOR | Jihun Kim, Javad Lavaei |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 104 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-07288