Optimistic Online Non-stochastic Control via FTRL
Not on the map. Why?
Abstract (arXiv)
This paper brings the concept of ``optimism" to the new and promising framework of online Non-stochastic Control (NSC). Namely, we study how NSC can benefit from a prediction oracle of unknown quality responsible for forecasting future costs. The posed problem is first reduced to an optimistic learning with delayed feedback problem, which is handled through the Optimistic Follow the Regularized Leader (OFTRL) algorithmic family. This reduction enables the design of textttOptFTRL-C, the first Disturbance Action Controller (DAC) with optimistic policy regret bounds. These new bounds are commensurate with the oracle's accuracy, ranging from O(1) for perfect predictions to the order-optimal O(√T) even when all predictions fail. By addressing the challenge of incorporating untrusted predictions into online control, this work contributes to the advancement of the NSC framework and paves the way toward effective and robust learning-based controllers.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 35 of its 51 references that are in the map. First posted on arXiv 4 Apr 2024.
Builds on
35 of 51 references.
References · 35
- 2019Introduction to Online Convex Optimizationcited by 793
- 2019Online Learning: A Modern Introduction Using Convex Optimizationcited by 407
- 2019Online Control with Adversarial Disturbancescited by 92
- 2020Improper Learning for Non-Stochastic Controlcited by 83
- 2018Online Linear Quadratic Controlcited by 55
- 2019The Nonstochastic Control Problemcited by 52
- 2020Logarithmic Regret for Adversarial Online Controlcited by 41
- 2019Online Optimal Control with Linear Dynamics and Predictions: Algorithms and Regret Analysiscited by 40
- 2019Logarithmic Regret for Online Controlcited by 38
- 2020Adaptive Regret for Control of Time-Varying Dynamicscited by 38
- 2021Non-stationary Online Learning with Memory and Non-stochastic Controlcited by 33
- 2021Online Learning with Optimism and Delaycited by 33
23 more
- 2020The Power of Predictions in Online Controlcited by 31
- 2021Perturbation-based Regret Analysis of Predictive Control in Linear Time Varying Systemscited by 25
- 2020Making Non-Stochastic Control (Almost) as Easy as Stochasticcited by 24
- 2021On the Regret Analysis of Online LQR Control with Predictionscited by 22
- 2020Online Optimal Control with Affine Constraintscited by 20
- 2020Non-Stochastic Control with Bandit Feedbackcited by 17
- 2022Introduction to Online Controlcited by 17
- 2015Online Convex Optimization Using Predictionscited by 16
- 2022Bounded-Regret MPC via Perturbation Analysis: Prediction Error, Constraints, and Nonlinearitycited by 16
- 2021Adversarial Tracking Control via Strongly Adaptive Online Learning with Memorycited by 14
- 2021Robustness and Consistency in Linear Quadratic Control with Untrusted Predictionscited by 14
- 2022Optimism in Face of a Context: Regret Guarantees for Stochastic Contextual MDPcited by 11
- 2022Optimistic No-regret Algorithms for Discrete Cachingcited by 10
- 2020Competitive Control with Delayed Imperfect Informationcited by 8
- 2021Competitive Controlcited by 8
- 2023Online Submodular Maximization via Online Convex Optimizationcited by 6
- 2022On the Regret of H∞ Controlcited by 5
- 2022Best of Both Worlds in Online Control: Competitive Ratio and Policy Regretcited by 5
- 2023Efficient Online Learning with Memory via Frank-Wolfe Optimization: Algorithms with Bounded Dynamic Regret and Applications to Controlcited by 5
- 2023Online Nonstochastic Control with Adversarial and Static Constraintscited by 5
- 2022Online Caching with no Regret: Optimistic Learning via Recommendationscited by 4
- 2024Fair Resource Allocation in Virtualized O-RAN Platformscited by 4
- 2023Adaptive Online Non-stochastic Controlcited by 3
Built on by
Citing papers per year: 2025: 2; 2026: 1.
3 papers.