Local kernels and the geometric structure of data
From MaRDI portal
Abstract: We introduce a theory of local kernels, which generalize the kernels used in the standard diffusion maps construction of nonparametric modeling. We prove that evaluating a local kernel on a data set gives a discrete representation of the generator of a continuous Markov process, which converges in the limit of large data. We explicitly connect the drift and diffusion coefficients of the process to the moments of the kernel. Moreover, when the kernel is symmetric, the generator is the Laplace-Beltrami operator with respect to a geometry which is influenced by the embedding geometry and the properties of the kernel. In particular, this allows us to generate any Riemannian geometry by an appropriate choice of local kernel. In this way, we continue a program of Belkin, Niyogi, Coifman and others to reinterpret the current diverse collection of kernel-based data analysis methods and place them in a geometric framework. We show how to use this framework to design local kernels invariant to various features of data. These data-driven local kernels can be used to construct conformally invariant embeddings and reconstruct global diffeomorphisms.
Recommendations
- Localized algorithms for multiple kernel learning
- scientific article; zbMATH DE number 2089352
- Kernels and distances for structured data
- The geometry of kernelized spectral clustering
- Kernel class-wise locality preserving projection
- Geometric structures arising from kernel density estimation on Riemannian manifolds
- Connecting dots: from local covariance to empirical intrinsic geometry and locally linear embedding
- A local linear embedding algorithm based on the harmonic-mean geodesic kernel
- Geometric perspectives on reproducing kernels
Cites work
- scientific article; zbMATH DE number 3810550 (Why is no real title available?)
- scientific article; zbMATH DE number 1201579 (Why is no real title available?)
- Anisotropic diffusion on sub-manifolds with application to Earth structure classification
- Diffusion Maps, Reduction Coordinates, and Low Dimensional Representation of Stochastic Systems
- Diffusion maps
- Diffusion maps for changing data
- Diffusion maps, spectral clustering and reaction coordinates of dynamical systems
- Embedology
- From graph to manifold Laplacian: the convergence rate
- Laplacian Eigenmaps for Dimensionality Reduction and Data Representation
- Learning Theory
- Machine Learning: ECML 2004
- Manifold Learning for Latent Variable Inference in Dynamical Systems
- Non-linear independent component analysis with diffusion maps
- Parametrization of Linear Systems Using Diffusion Kernels
- Regularization on graphs with function-adapted diffusion processes
- Riemannian geometry and geometric analysis
- The Laplacian on a Riemannian manifold. An introduction to analysis on manifolds
- Time-scale separation from diffusion-mapped delay coordinates
- Variable bandwidth diffusion kernels
- Vector diffusion maps and the connection Laplacian
Cited in
(42)- Density estimation on manifolds with boundary
- Delay-coordinate maps and the spectra of Koopman operators
- Data-driven spectral decomposition and forecasting of ergodic dynamical systems
- Kernel Methods for Bayesian Elliptic Inverse Problems on Manifolds
- Discovering causal structure with reproducing-kernel Hilbert space \(\epsilon\)-machines
- A geometrically constrained manifold embedding for an extrinsic Gaussian process
- Geometric triangulations and discrete Laplacians on manifolds: an update
- Delay-coordinate maps, coherence, and approximate spectra of evolution operators
- Diffusion-based kernel methods on Euclidean metric measure spaces
- Variable bandwidth diffusion kernels
- Multi-view kernel consensus for data analysis
- A Geometric Approach to the Transport of Discontinuous Densities
- Iterated diffusion maps for feature identification
- Dynamics-adapted cone kernels
- Operator-theoretic framework for forecasting nonlinear time series with kernel analog techniques
- A geometric heat-flow theory of Lagrangian coherent structures
- Local regularization of noisy point clouds: improved global geometric estimates and data analysis
- Diffusion representation for asymmetric kernels
- Reproducing kernel Hilbert space compactification of unitary evolution groups
- Diffusion maps tailored to arbitrary non-degenerate Itô processes
- Spatiotemporal pattern extraction by spectral analysis of vector-valued observables
- The diffusion geometry of fibre bundles: horizontal diffusion maps
- Diffusion maps for embedded manifolds with boundary with applications to PDEs
- Approximating solutions of linear elliptic PDE's on a smooth manifold using local kernel
- Extraction and prediction of coherent patterns in incompressible flows through space-time koopman analysis
- Ensemble forecasts in reproducing kernel Hilbert space family
- Time coupled diffusion maps
- Manifold learning with bi-stochastic kernels
- Fractional diffusion maps
- Koopman spectra in reproducing kernel Hilbert spaces
- Kernel-based methods for solving time-dependent advection-diffusion equations on manifolds
- Conditional expectation using compactification operators
- A spectral series approach to high-dimensional nonparametric regression
- Computing committors in collective variables via Mahalanobis diffusion maps
- A survey of statistical learning techniques as applied to inexpensive pediatric obstructive sleep apnea data
- Ghost Point Diffusion Maps for Solving Elliptic PDEs on Manifolds with Classical Boundary Conditions
- Koopman analysis of the long-term evolution in a turbulent convection cell
- Balancing geometry and density: path distances on high-dimensional data
- Manifold learning for organizing unstructured sets of process observations
- Spectral Discovery of Jointly Smooth Features for Multimodal Data
- Local conformal autoencoder for standardized data coordinates
- Nonparametric uncertainty quantification for stochastic gradient flows
This page was built for publication: Local kernels and the geometric structure of data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q265690)
