Data-Driven Kernel Designs for Optimized Greedy Schemes: A Machine Learning Perspective
From MaRDI portal
Abstract: Thanks to their easy implementation via Radial Basis Functions (RBFs), meshfree kernel methods have been proved to be an effective tool for e.g. scattered data interpolation, PDE collocation, classification and regression tasks. Their accuracy might depend on a length scale hyperparameter, which is often tuned via cross validation schemes. Here we leverage approaches and tools from the machine learning community to introduce two-layered kernel machines, which generalize the classical RBF approaches that rely on a single hyperparameter. Indeed, the proposed learning strategy returns a kernel that is optimized not only in the Euclidean directions, but that further incorporates kernel rotations. The kernel optimization is shown to be robust by using recently improved calculations of cross validation scores. Finally, the use of greedy approaches, and specifically of the Vectorial Kernel Orthogonal Greedy Algorithm (VKOGA), allows us to construct an optimized basis that adapts to the data. Beyond a rigorous analysis on the convergence of the so-constructed two-Layered (2L)-VKOGA, its benefits are highlighted on both synthesized and real benchmark data sets.
Recommendations
- Learning Theory
- A novel class of stabilized greedy kernel approximation algorithms: convergence, stability and uniform point distribution
- Biomechanical surrogate modelling using stabilized vectorial greedy kernel methods
- Learning the kernel function via regularization
- On the convergence of generalized kernel-based interpolation by greedy data selection algorithms
Cites work
- scientific article; zbMATH DE number 45848 (Why is no real title available?)
- A Correspondence Between Bayesian Estimation on Stochastic Processes and Smoothing by Splines
- A greedy non-intrusive reduced order model for shallow water equations
- A new variable shape parameter strategy for RBF approximation using neural networks
- A novel class of stabilized greedy kernel approximation algorithms: convergence, stability and uniform point distribution
- A representer theorem for deep kernel learning
- A stochastic extended Rippa's algorithm for LpOCV
- Active subspace methods in theory and practice: applications to kriging surfaces
- Active subspaces. Emerging ideas for dimension reduction in parameter studies
- Adaptive LOOCV-based kernel methods for solving time-dependent BVPs
- Adaptive greedy techniques for approximate solution of large RBF systems
- An algorithm for selecting a good value for the parameter \(c\) in radial basis function interpolation
- Analysis of target data-dependent greedy kernel algorithms: convergence rates for \(f\)-, \(f \cdot P\)- and \(f/P\)-greedy
- Convergence rate of the data-independent \(P\)-greedy algorithm in kernel-based approximation
- Deep learning
- Efficient reduced basis algorithm (ERBA) for kernel-based approximation
- Feature augmentation for the inversion of the Fourier transform with limited data
- Generalized Cross-Validation as a Method for Choosing a Good Ridge Parameter
- Interpolation in the limit of increasingly flat radial basis functions
- Jumping with variably scaled discontinuous kernels (VSDKs)
- Kernel flows: from learning kernels from data into the abyss
- Learning dynamical systems from data: a simple cross-validation perspective. I: Parametric kernel flows
- Learning via variably scaled kernels
- Meshfree approximation methods with Matlab. With CD-ROM.
- Near-optimal data-independent point locations for radial basis function interpolation
- On the search of the shape parameter in radial basis functions using univariate global optimization methods
- Principal angles between subspaces in an A-based scalar product: Algorithms and perturbation estimates
- Scattered Data Approximation
- Stable computation of multiquadric interpolants for all values of the shape parameter
- Surrogate modeling of multiscale models using kernel methods
- The Runge phenomenon and spatially variable shape parameters in RBF interpolation
- The extension of Rippa's algorithm beyond LOOCV
Cited in
(7)- A recipe for learning variably scaled kernels via discontinuous neural networks
- Classifier-dependent feature selection via greedy methods
- Application of deep kernel models for certified and adaptive RB-ML-ROM surrogate modeling
- Multi-fidelity learning of reduced order models for parabolic PDE constrained optimization
- Learning a robust shape parameter for RBF approximation
- Multiscale scattered data analysis in samplet coordinates
- A comparison study of supervised learning techniques for the approximation of high dimensional functions and feedback control
This page was built for publication: Data-Driven Kernel Designs for Optimized Greedy Schemes: A Machine Learning Perspective
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6154961)
