About me
I’m a Presidential and CAIRFI (Center for AI and Responsible Financial Innovation) fellow and a third-year Ph.D. student in the Department of Electrical Engineering at Columbia University advised by Prof. James Anderson. Before joining Columbia, I was an undergraduate research assistant in the Department of Engineering Science at the University of Oxford. I was awarded my M.S. in Control, Signal, and Image Processing from the University of Paris-Saclay and the M.Eng. in Electrical Engineering from CentraleSupélec both in 2022. In addition, I received my B.Eng. in Electrical Engineering from the University of Campinas (Unicamp) also in 2022.
Research Interests
My research interests lie at the intersection of control theory, machine learning, and optimization.
Collaborators
Over the past year, I had the pleasure of collaborating with brilliant people:
Students: Han Wang (FL for estimation and control), Thomas Zhang (multi-task rep. learning), Bruce D. Lee (adaptive control), Donglin Zhan (meta-learning LQR).
Professors: James Anderson (Columbia University), Nikolai Matni (UPenn), Aritra Mitra (NC State University).
Papers
[10] LF. Toso* , H. Wang*, J. Anderson, “Asynchronous Heterogeneous Linear Quadratic Regulator Design”. To appear at CDC, 2024.
[9] Bruce Lee, LF. Toso* , T. Zhang, J. Anderson, N. Matni, “Regret Analysis of Multi-task Representation Learning for Linear-Quadratic Adaptive Control”. Under review, 2024.
[8] Best Paper Award - LF. Toso, D. Zhang, J. Anderson, H. Wang, “Meta-Learning Linear Quadratic Regulators: A Policy Gradient MAML Approach for Model-free LQR”. L4DC, 2024 (oral - top 7.5%).
[7] T. Zhang, LF. Toso, J. Anderson, N. Matni, “Sample-Efficient Linear Representation Learning from Non-IID Non-Isotropic Data”. ICLR, 2024 (spotlight - top 5%).
[6] LF. Toso, H. Wang, J. Anderson, “Oracle Complexity Reduction for Model-free LQR: A Stochastic Variance-Reduced Policy Gradient Approach”. ACC, 2024.
[5] H. Wang, LF. Toso, A. Mitra, J. Anderson, “Model-free Learning with Heterogeneous Dynamical Systems: A Federated LQR Approach”. Under review, 2023.
[4] LF. Toso, H. Wang, J. Anderson, “Learning Personalized Models with Clustered System Identification”. CDC, 2023.
[3] H. Wang, LF. Toso, J. Anderson, “Fedsysid: A federated approach to sample-efficient system identification”. L4DC, 2023.
[2] LF. Toso, R. Drummond, S. Duncan, “Regional stability analysis of transitional fluid flows”. IEEE Control Systems Letters, 2022.
[1] LF. Toso, G. Valmorbida, “Lyapunov Function computation for Periodic Linear Hybrid Systems via Handelman, Polya and SoS approaches: A comparative study”. IFAC CAO, 2022.
^* denotes equal contribution.
Talks
ACC, 2024, Toronto, Canada. “Oracle Complexity Reduction for Model-free LQR: A Stochastic Variance-Reduced Policy Gradient Approach”. Slides.
L4DC, 2024, Oxford, UK. “Meta-Learning Linear Quadratic Regulators: A Policy Gradient MAML Approach for Model-free LQR”. Slides.
Mentoring
Over the Summer and Fall of 2023, I had the pleasure of mentoring two outstanding M.S. EE students at Columbia:
- Patrick Munar on “Controlling Balanced Intravenous and Volatile Anesthesia” (paper in preparation).
- Zhe Mo on “SOSPy: A Python library for solving sum-of-squares programming”.
Teaching
I love teaching! Over the past few years I had the chance of serving as a head TA for:
- Convex Optimization, Fall 2023, Columbia University.
- Modern Control Theory, Spring 2023, Columbia University.
- Convex Optimization, Fall 2022, Columbia University.
- Control Theory, Spring 2022, University of Campinas.
- Circuit Analysis I, Fall 2018, University of Campinas.
- Calculus I, Fall 2017, University of Campinas.
Peer Reviewer
- IEEE Transactions on Automatic Control (TAC).
- IEEE Control Systems Letters (L-CSS).
- IEEE Transactions on Vehicular Technology (TVT).
- Learning for Dynamics & Control Conference (L4DC).
- IEEE Conference on Decision and Control (CDC).
- American Control Conference (ACC).
- International Federation of Automatic Control (IFAC).