Deterministic error bounds for kernel-based learning techniques under bounded noise
From MaRDI portal
Abstract: We consider the problem of reconstructing a function from a finite set of noise-corrupted samples. Two kernel algorithms are analyzed, namely kernel ridge regression and -support vector regression. By assuming the ground-truth function belongs to the reproducing kernel Hilbert space of the chosen kernel, and the measurement noise affecting the dataset is bounded, we adopt an approximation theory viewpoint to establish extit{deterministic}, finite-sample error bounds for the two models. Finally, we discuss their connection with Gaussian processes and two numerical examples are provided. In establishing our inequalities, we hope to help bring the fields of non-parametric kernel learning and system identification for robust control closer to each other.
Recommendations
- Deterministic error analysis of support vector regression and related regularized kernel methods
- Learning a function from noisy samples at a finite sparse set of points
- Learning with generalization capability by kernel methods of bounded complexity
- Learning theory estimates via integral operators and their approximations
- Error bounds for learning the kernel
Cites work
- A new kernel-based approach for linear system identification
- A New Kernel-Based Approach for NonlinearSystem Identification
- A nonparametric kernel-based approach to Hammerstein system identification
- A shift in paradigm for system identification
- A unified theory of radial basis functions. Native Hilbert spaces for radial basis functions. II
- Bounded noises in physics, biology, and engineering
- Concentration estimates for learning with unbounded sampling
- Continuous-Time DC Kernel—A Stable Generalized First-Order Spline Kernel
- Feedback Linearization Based on Gaussian Processes With Event-Triggered Online Learning
- Gaussian processes for machine learning.
- High-dimensional Bayesian optimization with projections using quantile Gaussian processes
- scientific article; zbMATH DE number 1804115 (Why is no real title available?)
- scientific article; zbMATH DE number 2217387 (Why is no real title available?)
- Improved estimates for condition numbers of radial basis function interpolation matrices
- Information-Theoretic Regret Bounds for Gaussian Process Optimization in the Bandit Setting
- Kernel Approximation on Manifolds I: Bounding the Lebesgue Constant
- Kernel methods in system identification, machine learning and function estimation: a survey
- Kernel-based identification of non-causal systems with application to inverse model control
- Learning rates of least-square regularized regression
- Learning theory estimates via integral operators and their approximations
- Maximum Entropy Kernels for System Identification
- Near-optimal data-independent point locations for radial basis function interpolation
- Optimal learning rates for least squares regularized regression with unbounded sampling
- Robust learning-based MPC for nonlinear constrained systems
- Theory of Reproducing Kernels
- Transfer function and transient estimation by Gaussian process regression in the frequency domain
- Universal kernels
Cited in
(12)- Learning a function from noisy samples at a finite sparse set of points
- High-probability stable Gaussian process-supported model predictive control for Lur'e systems
- Bayesian frequentist bounds for machine learning and system identification
- Deterministic error analysis of support vector regression and related regularized kernel methods
- Learning stability guarantees for constrained switching linear systems from noisy observations
- Lower bounds on the noiseless worst-case complexity of efficient global optimization
- Error bounds and the asymptotic setting in kernel-based approximation
- Decentralized event-triggered online learning for safe consensus control of multi-agent systems with Gaussian process regression
- L^-error bounds for approximations of the Koopman operator by kernel extended dynamic mode decomposition
- Polynomial-based kernel reproduced gradient descent for stochastic optimization
- Robust data-driven tube mpc for nonlinear systems with deterministic kernel error bounds
- Deterministic error analysis for learning from non-point-evaluation functional data
This page was built for publication: Deterministic error bounds for kernel-based learning techniques under bounded noise
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2665700)
