Research map / Learning theory and algorithms / Sample complexity
Near optimal finite time identification of arbitrary linear 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 |
|---|---|---|---|
| Use a coupled analysis of sample covariance and the self-normalized martingale term | OLS system identification · Stable and marginally stable systems | OLS estimation error: √(C₀(d) log(1/δ) / T); C₁(d)/T log(T/δ) | Helped |
| Use anti-concentration arguments and subgaussian tail inequalities for explosive systems | OLS system identification · Regular explosive systems | OLS estimation error: C₂(d) σmax(A⁻ᵀ)/δ | Helped |
| Extend the analysis to systems with eigenvalues in stable, marginally stable, and explosive regimes | OLS system identification · General regular systems | OLS estimation error: poly(log T, log(1/δ))/T; poly(log T, log(1/δ))/√T | Helped |
| Use OLS without the regularity condition on explosive eigenvalues | OLS system identification · Irregular explosive systems | Consistency | Hurt |
| Rely on signal-to-noise ratio rather than sample-covariance conditioning to predict consistency | OLS system identification · Explosive systems | Consistency | Hurt |
Closest papers by meaning
- Finite Sample Analyses for Continuous-time Linear Systems: System Identification and Online Control (NeurIPS 2025 · similarity 0.88)
- Stabilizing LTI Systems under Partial Observability: Sample Complexity and Fundamental Limits (NeurIPS 2025 · similarity 0.86)
- Identification of Analytic Nonlinear Dynamical Systems with Non-asymptotic Guarantees (NeurIPS 2024 · similarity 0.85)
- A Memory Efficient Unified Algorithm for Online Learning of Linear Dynamical Systems (NeurIPS 2026 · similarity 0.83)
- PAC-Bayesian Error Bound, via Rényi Divergence, for a Class of Linear Time-Invariant State-Space Models (ICML 2024 · similarity 0.83)
- Consistent Long-Term Forecasting of Ergodic Dynamical Systems (ICML 2024 · similarity 0.82)
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.