System Identification & Learning Dynamical Systems Reading List

Curated by Mouhssine Rifaki | Stanford Electrical Engineering | Last updated August 2026

Learning to predict and control dynamical systems sits at the intersection of control theory and machine learning. These papers span classical identification, online learning for linear systems, and modern spectral and deep approaches to forecasting.

System Identification & Learning Dynamical Systems: 10 key papers

1. Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification
   Simchowitz et al. arXiv 2018.
2. Finite Time Identification in Unstable Linear Systems
   Faradonbeh et al. arXiv 2017.
3. Non-asymptotic Identification of LTI Systems from a Single Trajectory
   Oymak and Ozay. arXiv 2018.
4. Finite Time LTI System Identification
   Sarkar et al. Journal of Machine Learning Research 2021.
5. Discovering governing equations from data by sparse identification of nonlinear dynamical systems
   Brunton et al. Proceedings of the National Academy of Sciences 2016.
6. Data-driven discovery of coordinates and governing equations
   Champion et al. arXiv 2019.
7. Deep learning for universal linear embeddings of nonlinear dynamics
   Lusch et al. arXiv 2017.
8. Learning nonlinear dynamical systems from a single trajectory
   Foster et al. arXiv 2020.
9. System Identification via Nuclear Norm Regularization
   Sun et al. arXiv 2022.
10. Statistical Learning Theory for Control: A Finite Sample Perspective
    Tsiamis et al. arXiv 2022.

Related reading lists

• Spectral Methods in Machine Learning
• Multi-Armed Bandits & Best-Arm Identification
• Optimal Control & LQR/LQG
• Time Series Forecasting & Temporal Models
• Low-Rank Structure in Machine Learning
• Neuroscience-Inspired Reinforcement Learning
• Physics-Informed Machine Learning & Scientific ML
• Game Theory for Multi-Agent Systems