Research map / Learning theory and algorithms / Kernels and regression theory
Nonparametric adaptive control and prediction: theory and randomized algorithms
Nicholas M. Boffi, Stephen Tu, Jean-Jacques E. Slotine
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| increase the number of random Fourier features K in the adaptive control input (ablation) | Randomized adaptive controller · 5-dim synthetic adaptive control | tracking error | Improved |
| use the exact nonparametric kernel input instead of its random feature approximation (ablation) | Randomized adaptive controller · 5-dim synthetic adaptive control | tracking and interpolation error: best performance by several orders of magnitude | Improved |
| increase K in the random feature adaptive input (ablation) | Randomized adaptive controller · 5-dim synthetic adaptive control | input magnitude | Improved |
| increase number of features K (ablation) | Randomized adaptive predictor · m-body system, 60-dim | asymptotic prediction error: power law K^-xi with xi ≈ 1.28 ± 0.03 | Improved |
| increase number of features K (ablation) | Randomized adaptive predictor · m-body system, 60-dim | interpolation error: power law K^-zeta with zeta ≈ 0.77 ± 0.03 | Improved |
| add online learning of the dynamics to the predictor (ablation) | Randomized adaptive predictor · m-body system, 60-dim | position prediction accuracy | Improved |
| replace random features with a single hidden-layer neural network adaptation law | Adaptive controller, unstable 5-dim system · Unstable system with time-varying trajectory | tracking error: slightly improved performance for widths 32 and 64 | Improved |
| replace random features with neural network adaptation law | Adaptive controller, unstable 5-dim system · Unstable system with time-varying trajectory | stability across learning rates | Hurt |
| replace random features with neural network adaptation law | Adaptive controller, unstable 5-dim system · Unstable system with time-varying trajectory | initial input norm: one or two orders of magnitude larger input | Hurt |
Closest papers by meaning
- Kernel-based guarantees for nonlinear parametric models in Bayesian optimization (NeurIPS 2026 · similarity 0.86)
- Safely Learning Controlled Stochastic Dynamics (NeurIPS 2025 · similarity 0.86)
- Memory-Augmented Potential Field Theory: A Framework for Adaptive Control in Non-Convex Domains (NeurIPS 2025 · similarity 0.85)
- Optimal Control Operator Perspective and a Neural Adaptive Spectral Method (AAAI 2025 · similarity 0.85)
- A Memory Efficient Unified Algorithm for Online Learning of Linear Dynamical Systems (NeurIPS 2026 · similarity 0.85)
- FESSNC: Fast Exponentially Stable and Safe Neural Controller (ICML 2024 · similarity 0.85)
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.
