Research map / Bandits and online learning / Online convex optimization and games
Online Control of Unknown Time-Varying Dynamical Systems
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Allow unknown time-varying dynamics with nonzero system variability | Online control against DRC or DAC policies · Unknown LTV systems | expected regret: C \cdot T \sigma | Hurt |
| Use adaptive exploration to estimate changing Markov operators, then apply DRC-OGD | Unknown LTV control against DRC policies · Any interval I | adaptive regret: |I| \sqrt{\mathbb{E}[\mathrm{Var}_{I}(\mathbf{G})]} + d_u T^{2/3} | Helped |
| Set system variability to zero | Ada-Ctrl on unknown LTI systems · Unknown LTI systems | adaptive regret: T^{2/3} | Helped |
| Compete with state-feedback policies using an inefficient discretized Exp3 bandit method | State-feedback control · Stabilizing state-feedback controllers | expected regret: T^{1-\frac{1}{2(d_{x} d_{u} +3)}} | Helped |
| Require strongly adaptive regret with a polynomial-in-interval-length guarantee | Online prediction with costly noisy-oracle feedback · Partial-information vector-sequence prediction | strongly adaptive regret: \tilde{O}(|I|^{1-\gamma}) | Hurt |
| Replace LTI dynamics with LTV dynamics | Equivalence among DRC, DAC, and linear feedback policy classes · LTV systems | relative control performance: \Omega(T) | Mixed |
| Optimize offline over state-feedback controllers on sequentially stable LTV systems | State-feedback planning · Max-3Sat reduction | computational complexity: T = \Theta(mn) | Hurt |
Closest papers by meaning
- Naive Exploration is Optimal for Online LQR (ICML 2020 · similarity 0.92)
- Online Adaptive Policy Selection in Time-Varying Systems: No-Regret via Contractive Perturbations (NeurIPS 2023 · similarity 0.91)
- Regret Bounds for Episodic Risk-Sensitive Linear Quadratic Regulator (ICLR 2025 · similarity 0.89)
- Online Inverse Linear Optimization: Efficient Logarithmic-Regret Algorithm, Robustness to Suboptimality, and Lower Bound (NeurIPS 2025 · similarity 0.89)
- Online Control with Adversarial Disturbance for Continuous-time Linear Systems (NeurIPS 2024 · similarity 0.89)
- A Memory Efficient Unified Algorithm for Online Learning of Linear Dynamical Systems (NeurIPS 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.