Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Stephen Tu

Stephen Tu

Closing the Closed-Loop Distribution Shift in Safe Imitation Learning


Feb 18, 2021
Stephen Tu, Alexander Robey, Nikolai Matni

Figure 1 for Closing the Closed-Loop Distribution Shift in Safe Imitation Learning
Figure 2 for Closing the Closed-Loop Distribution Shift in Safe Imitation Learning
Figure 3 for Closing the Closed-Loop Distribution Shift in Safe Imitation Learning
Figure 4 for Closing the Closed-Loop Distribution Shift in Safe Imitation Learning

  Access Paper or Ask Questions

Learning Robust Hybrid Control Barrier Functions for Uncertain Systems


Jan 16, 2021
Alexander Robey, Lars Lindemann, Stephen Tu, Nikolai Matni

Figure 1 for Learning Robust Hybrid Control Barrier Functions for Uncertain Systems
Figure 2 for Learning Robust Hybrid Control Barrier Functions for Uncertain Systems
Figure 3 for Learning Robust Hybrid Control Barrier Functions for Uncertain Systems
Figure 4 for Learning Robust Hybrid Control Barrier Functions for Uncertain Systems
* 17 pages, 7th IFAC Conference on Analysis and Design of Hybrid Systems (submitted). arXiv admin note: text overlap with arXiv:2011.04112 
  Access Paper or Ask Questions

Regret Bounds for Adaptive Nonlinear Control


Nov 26, 2020
Nicholas M. Boffi, Stephen Tu, Jean-Jacques E. Slotine

Figure 1 for Regret Bounds for Adaptive Nonlinear Control

  Access Paper or Ask Questions

Safely Learning Dynamical Systems from Short Trajectories


Nov 24, 2020
Amir Ali Ahmadi, Abraar Chaudhry, Vikas Sindhwani, Stephen Tu

Figure 1 for Safely Learning Dynamical Systems from Short Trajectories
Figure 2 for Safely Learning Dynamical Systems from Short Trajectories
Figure 3 for Safely Learning Dynamical Systems from Short Trajectories

  Access Paper or Ask Questions

Learning Hybrid Control Barrier Functions from Data


Nov 08, 2020
Lars Lindemann, Haimin Hu, Alexander Robey, Hanwen Zhang, Dimos V. Dimarogonas, Stephen Tu, Nikolai Matni

Figure 1 for Learning Hybrid Control Barrier Functions from Data
Figure 2 for Learning Hybrid Control Barrier Functions from Data
Figure 3 for Learning Hybrid Control Barrier Functions from Data
Figure 4 for Learning Hybrid Control Barrier Functions from Data
* 27 pages, Conference on Robot Learning 2020 
  Access Paper or Ask Questions

Learning Stability Certificates from Data


Sep 14, 2020
Nicholas M. Boffi, Stephen Tu, Nikolai Matni, Jean-Jacques E. Slotine, Vikas Sindhwani

Figure 1 for Learning Stability Certificates from Data
Figure 2 for Learning Stability Certificates from Data
Figure 3 for Learning Stability Certificates from Data
* Fixes an error in the statement and proof of Theorem 5.1, Theorem 5.2, and Proposition D.1 
  Access Paper or Ask Questions

Learning Control Barrier Functions from Expert Demonstrations


Apr 07, 2020
Alexander Robey, Haimin Hu, Lars Lindemann, Hanwen Zhang, Dimos V. Dimarogonas, Stephen Tu, Nikolai Matni

Figure 1 for Learning Control Barrier Functions from Expert Demonstrations
Figure 2 for Learning Control Barrier Functions from Expert Demonstrations
Figure 3 for Learning Control Barrier Functions from Expert Demonstrations
Figure 4 for Learning Control Barrier Functions from Expert Demonstrations

  Access Paper or Ask Questions

Observational Overfitting in Reinforcement Learning


Dec 28, 2019
Xingyou Song, Yiding Jiang, Stephen Tu, Yilun Du, Behnam Neyshabur

Figure 1 for Observational Overfitting in Reinforcement Learning
Figure 2 for Observational Overfitting in Reinforcement Learning
Figure 3 for Observational Overfitting in Reinforcement Learning
Figure 4 for Observational Overfitting in Reinforcement Learning
* Published as a conference paper in ICLR 2020 
  Access Paper or Ask Questions

A Tutorial on Concentration Bounds for System Identification


Jun 27, 2019
Nikolai Matni, Stephen Tu

Figure 1 for A Tutorial on Concentration Bounds for System Identification
* Tutorial paper submitted to 2020 IEEE Conference on Decision and Control 
  Access Paper or Ask Questions

From self-tuning regulators to reinforcement learning and back again


Jun 27, 2019
Nikolai Matni, Alexandre Proutiere, Anders Rantzer, Stephen Tu

Figure 1 for From self-tuning regulators to reinforcement learning and back again
Figure 2 for From self-tuning regulators to reinforcement learning and back again
Figure 3 for From self-tuning regulators to reinforcement learning and back again
* Tutorial paper submitted to 2020 IEEE Conference on Decision and Control 
  Access Paper or Ask Questions

Finite-time Analysis of Approximate Policy Iteration for the Linear Quadratic Regulator


May 30, 2019
Karl Krauth, Stephen Tu, Benjamin Recht

Figure 1 for Finite-time Analysis of Approximate Policy Iteration for the Linear Quadratic Regulator
Figure 2 for Finite-time Analysis of Approximate Policy Iteration for the Linear Quadratic Regulator

  Access Paper or Ask Questions

Certainty Equivalent Control of LQR is Efficient


Feb 21, 2019
Horia Mania, Stephen Tu, Benjamin Recht

* 21 pages 
  Access Paper or Ask Questions

The Gap Between Model-Based and Model-Free Methods on the Linear Quadratic Regulator: An Asymptotic Viewpoint


Dec 09, 2018
Stephen Tu, Benjamin Recht


  Access Paper or Ask Questions

Minimax Lower Bounds for $\mathcal{H}_\infty$-Norm Estimation


Sep 28, 2018
Stephen Tu, Ross Boczar, Benjamin Recht

Figure 1 for Minimax Lower Bounds for $\mathcal{H}_\infty$-Norm Estimation
Figure 2 for Minimax Lower Bounds for $\mathcal{H}_\infty$-Norm Estimation

  Access Paper or Ask Questions

Safely Learning to Control the Constrained Linear Quadratic Regulator


Sep 26, 2018
Sarah Dean, Stephen Tu, Nikolai Matni, Benjamin Recht

Figure 1 for Safely Learning to Control the Constrained Linear Quadratic Regulator
Figure 2 for Safely Learning to Control the Constrained Linear Quadratic Regulator

  Access Paper or Ask Questions

Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification


May 24, 2018
Max Simchowitz, Horia Mania, Stephen Tu, Michael I. Jordan, Benjamin Recht


  Access Paper or Ask Questions

Regret Bounds for Robust Adaptive Control of the Linear Quadratic Regulator


May 23, 2018
Sarah Dean, Horia Mania, Nikolai Matni, Benjamin Recht, Stephen Tu

Figure 1 for Regret Bounds for Robust Adaptive Control of the Linear Quadratic Regulator
Figure 2 for Regret Bounds for Robust Adaptive Control of the Linear Quadratic Regulator

  Access Paper or Ask Questions

Learning Contracting Vector Fields For Stable Imitation Learning


Apr 13, 2018
Vikas Sindhwani, Stephen Tu, Mohi Khansari

Figure 1 for Learning Contracting Vector Fields For Stable Imitation Learning
Figure 2 for Learning Contracting Vector Fields For Stable Imitation Learning
Figure 3 for Learning Contracting Vector Fields For Stable Imitation Learning
Figure 4 for Learning Contracting Vector Fields For Stable Imitation Learning

  Access Paper or Ask Questions

On the Sample Complexity of the Linear Quadratic Regulator


Jan 25, 2018
Sarah Dean, Horia Mania, Nikolai Matni, Benjamin Recht, Stephen Tu

Figure 1 for On the Sample Complexity of the Linear Quadratic Regulator
Figure 2 for On the Sample Complexity of the Linear Quadratic Regulator
Figure 3 for On the Sample Complexity of the Linear Quadratic Regulator
Figure 4 for On the Sample Complexity of the Linear Quadratic Regulator
* Contains a new convex relaxation for the main optimization problem that is computationally inexpensive and performs well in simulation. The new relaxation uses a common Lyapunov function formulation for the mixed H2/H-infinity synthesis problem. We added new section on computational methods and new numerical experiments incorporating this technique. We also corrected typos and added references 
  Access Paper or Ask Questions

Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator


Dec 22, 2017
Stephen Tu, Benjamin Recht

Figure 1 for Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator
Figure 2 for Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator
Figure 3 for Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator
Figure 4 for Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator

  Access Paper or Ask Questions

Non-Asymptotic Analysis of Robust Control from Coarse-Grained Identification


Nov 30, 2017
Stephen Tu, Ross Boczar, Andrew Packard, Benjamin Recht

Figure 1 for Non-Asymptotic Analysis of Robust Control from Coarse-Grained Identification
Figure 2 for Non-Asymptotic Analysis of Robust Control from Coarse-Grained Identification
Figure 3 for Non-Asymptotic Analysis of Robust Control from Coarse-Grained Identification
Figure 4 for Non-Asymptotic Analysis of Robust Control from Coarse-Grained Identification
* A substantial revision, where we strengthen our existing upper bounds and introduce a matching lower bound 
  Access Paper or Ask Questions

CYCLADES: Conflict-free Asynchronous Machine Learning


May 31, 2016
Xinghao Pan, Maximilian Lam, Stephen Tu, Dimitris Papailiopoulos, Ce Zhang, Michael I. Jordan, Kannan Ramchandran, Chris Re, Benjamin Recht

Figure 1 for CYCLADES: Conflict-free Asynchronous Machine Learning
Figure 2 for CYCLADES: Conflict-free Asynchronous Machine Learning
Figure 3 for CYCLADES: Conflict-free Asynchronous Machine Learning
Figure 4 for CYCLADES: Conflict-free Asynchronous Machine Learning

  Access Paper or Ask Questions

Large Scale Kernel Learning using Block Coordinate Descent


Feb 17, 2016
Stephen Tu, Rebecca Roelofs, Shivaram Venkataraman, Benjamin Recht

Figure 1 for Large Scale Kernel Learning using Block Coordinate Descent
Figure 2 for Large Scale Kernel Learning using Block Coordinate Descent
Figure 3 for Large Scale Kernel Learning using Block Coordinate Descent
Figure 4 for Large Scale Kernel Learning using Block Coordinate Descent

  Access Paper or Ask Questions