![[2602.13485] Federated Learning of Nonlinear Temporal Dynamics with Graph Attention-based Cross-Client Interpretability](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.13485] Federated Learning of Nonlinear Temporal Dynamics with Graph Attention-based Cross-Client Interpretability
Summary
This paper presents a federated learning framework for understanding nonlinear temporal dynamics across decentralized systems, enhancing interpretability and privacy.
Why It Matters
As industrial systems increasingly rely on distributed sensors, understanding interdependencies in time series data becomes crucial. This research addresses the challenges of decentralized data sharing while maintaining model integrity and privacy, making it relevant for industries utilizing federated learning in complex environments.
Key Takeaways
- Introduces a federated framework for learning temporal interdependencies in decentralized systems.
- Utilizes Graph Attention Networks for improved interpretability of nonlinear dynamics.
- Demonstrates theoretical convergence guarantees and practical performance through synthetic and real-world experiments.
Computer Science > Machine Learning arXiv:2602.13485 (cs) [Submitted on 13 Feb 2026] Title:Federated Learning of Nonlinear Temporal Dynamics with Graph Attention-based Cross-Client Interpretability Authors:Ayse Tursucular, Ayush Mohanty, Nazal Mohamed, Nagi Gebraeel View a PDF of the paper titled Federated Learning of Nonlinear Temporal Dynamics with Graph Attention-based Cross-Client Interpretability, by Ayse Tursucular and 3 other authors View PDF HTML (experimental) Abstract:Networks of modern industrial systems are increasingly monitored by distributed sensors, where each system comprises multiple subsystems generating high dimensional time series data. These subsystems are often interdependent, making it important to understand how temporal patterns at one subsystem relate to others. This is challenging in decentralized settings where raw measurements cannot be shared and client observations are heterogeneous. In practical deployments each subsystem (client) operates a fixed proprietary model that cannot be modified or retrained, limiting existing approaches. Nonlinear dynamics further make cross client temporal interdependencies difficult to interpret because they are embedded in nonlinear state transition functions. We present a federated framework for learning temporal interdependencies across clients under these constraints. Each client maps high dimensional local observations to low dimensional latent states using a nonlinear state space model. A central server l...
