Research map / Reinforcement learning / Policy gradient and actor-critic
Learning convex bounds for linear quadratic control policy synthesis
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 |
|---|---|---|---|
| Use sequential SDP refinement instead of stopping at the common-Lyapunov relaxation (ablation) | LQR policy synthesis · Synthetic systems, LQR suboptimality | LQR suboptimality | Helped |
| Use stochastic sampled-model stability constraints rather than robust stability constraints | LQR policy synthesis · Synthetic systems, low-data regime | LQR cost on the true system | Helped |
| Apply the proposed policy synthesis method with 100 posterior samples | LQR policy synthesis · 95% confidence region, n_x=12 | median % unstable closed-loop models: 0.27 | Helped |
| Use robust methods instead of the proposed sampled-model method | LQR policy synthesis · Synthetic systems, n_x=12 | policy synthesis infeasibility: 46.0 | Hurt |
| Use the proposed method rather than least-squares policy iteration | Rotary inverted pendulum · Real-world rotary inverted pendulum | stabilization | Helped |
| Use the proposed sequential optimization method rather than the H2/H-infinity solution | H2/H-infinity initialization · Synthetic systems, LQR suboptimality | LQR suboptimality | Helped |
Closest papers by meaning
- Robust exploration in linear quadratic reinforcement learning (NeurIPS 2019 · similarity 0.92)
- Online Policy Gradient for Model Free Learning of Linear Quadratic Regulators with $\sqrt$T Regret (ICML 2021 · similarity 0.90)
- Learning Linear-Quadratic Regulators Efficiently with only $\sqrtT$ Regret (ICML 2019 · similarity 0.89)
- Robust Reinforcement Learning: A Case Study in Linear Quadratic Regulation (AAAI 2021 · similarity 0.89)
- Regret Bounds for Episodic Risk-Sensitive Linear Quadratic Regulator (ICLR 2025 · similarity 0.87)
- Certifying Stability of Reinforcement Learning Policies using Generalized Lyapunov Functions (NeurIPS 2025 · similarity 0.86)
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.