arXiv research

We construct radial fundamental solutions for the differential form Laplacian on negatively curved symmetric spaces. At least one of these Green's functions also yields a Biot-Savart Opearator, i.e. a right inverse of the exterior differential on closed forms with image in the kernel of the codifferential. Any Biot-Sav…

The paper connects linking numbers to Biot-Savart kernels on compact manifolds.

We study steady-state magnetic fields in the geometric setting of positive curvature on subdomains of the three-dimensional sphere. By generalizing the Biot-Savart law to an integral operator BS acting on all vector fields, we show that electrodynamics in such a setting behaves rather similarly to Euclidean electrodyna…

We introduce here explicit integral formulas for linking, twisting, writhing and helicity on the 3-sphere and in hyperbolic 3-space. These formulas, like their prototypes in Euclidean 3-space, are geometric rather than just topological, in the sense that their integrands are invariant under orientation-preserving isome…

Neural Networks improve incompressible flow simulations without complex kernels.

Extends Arnold's linking theory to higher dimensions and submanifolds.

We introduce and illustrate a new approach to the unknotting problem via the dynamics of vortex strings in a nonlinear partial differential equation of reaction-diffusion type. To untangle a given knot, a Biot-Savart construction is used to initialize the knot as a vortex string in the FitzHugh-Nagumo equation. Remarka…

New constraints rule out some optimal domains for helicity maximisation.

In this first of two papers, we develop a steady-state version of classical electrodynamics on the 3-sphere and in hyperbolic 3-space, including an explicit formula for the vector-valued Green's operator, an explicit formula of Biot-Savart type for the magnetic field, and a corresponding Ampere's Law contained in Maxwe…

We consider the $D$-dimensional Euclidean space, $\mathbb{R}^D$, with certain $(D-N)$-dimensional compact, closed and orientable sub-manifolds (which we call \emph{singularity manifolds} and represent by $\widetilde{\mathcal{S}}$) removed from it. We define and investigate the problem of finding a homotopy-like class i…

The paper introduces surfaces with constant solid angle for designing shell structures.

Cauchy invariants are now viewed as a powerful tool for investigating the Lagrangian structure of three-dimensional (3D) ideal flow (Frisch & Zheligovsky, Commun. Math. Phys., vol. 326, 2014, pp. 499-505, Podvigina et al., J. Comput. Phys., vol. 306, 2016, pp. 320-342). Looking at such invariants with the modern tools …

The helicity of a vector field is a measure of the average linking of pairs of integral curves of the field. Computed by a six-dimensional integral, it is widely useful in the physics of fluids. For a divergence-free field tangent to the boundary of a domain in 3-space, helicity is known to be invariant under volume-pr…

Many unsupervised kernel methods rely on the estimation of the kernel covariance operator (kernel CO) or kernel cross-covariance operator (kernel CCO). Both kernel CO and kernel CCO are sensitive to contaminated data, even when bounded positive definite kernels are used. To the best of our knowledge, there are few well…

Survey of kernels, RKHS, and their applications in machine learning.

To the best of our knowledge, there are no general well-founded robust methods for statistical unsupervised learning. Most of the unsupervised methods explicitly or implicitly depend on the kernel covariance operator (kernel CO) or kernel cross-covariance operator (kernel CCO). They are sensitive to contaminated data, …

Deep neural kernels and Laplace kernel have equivalent RKHS on spheres.

Kernel methods linked to feature subspaces and maximal correlation kernels.

PGF kernels analyze spherical data using generalized RBF kernels.

Adapts manifold structure for better clustering performance.

Optimal kernel in KR can be data-dependent, improving model performance.

Quantum kernels can be efficiently embedded into classical feature spaces.

New random feature maps for Laplacian and related kernels.

New method for learning with non-Euclidean data using decomposable kernels.

New estimator reduces kernel mean estimation error.

Estimates kernel eigenvalues for compositional dot-product kernels.

Optimal Biweight kernel and computationally efficient Epanechnikov kernel for modal linear regression.

We present Random Partition Kernels, a new class of kernels derived by demonstrating a natural connection between random partitions of objects and kernels between those objects. We show how the construction can be used to create kernels from methods that would not normally be viewed as random partitions, such as Random…

The NNGP kernel's predictions closely match those of the Matern kernel under certain conditions.

In this paper, we compare 5 different nonlinear kernels: min-max, RBF, fRBF (folded RBF), acos, and acos-$χ^2$, on a wide range of publicly available datasets. The proposed fRBF kernel performs very similarly to the RBF kernel. Both RBF and fRBF kernels require an important tuning parameter ($γ$). Interestingly, for a …

New kernels allow learning from non-separable data.

Laplace kernel and Neural Tangent Kernels are shown to be nearly identical for normalized data.

Kernel smoothing on unknown manifolds with bounds and asymptotic normality.

In this paper we propose a family of tractable kernels that is dense in the family of bounded positive semi-definite functions (i.e. can approximate any bounded kernel with arbitrary precision). We start by discussing the case of stationary kernels, and propose a family of spectral kernels that extends existing approac…

The success of kernel-based learning methods depend on the choice of kernel. Recently, kernel learning methods have been proposed that use data to select the most appropriate kernel, usually by combining a set of base kernels. We introduce a new algorithm for kernel learning that combines a {\em continuous set of base …

New kernels capture both local and non-local interactions efficiently.

Quantum kernel machines need to use more complex kernels to fully exploit their potential.

Constructing the adjacency graph is fundamental to graph-based clustering. Graph learning in kernel space has shown impressive performance on a number of benchmark data sets. However, its performance is largely determined by the chosen kernel matrix. To address this issue, the previous multiple kernel learning algorith…

Optimal kernel improves estimation accuracy in modal statistical methods.

Efficiently searches through Gaussian process kernels using symbolic representation and Bayesian optimization.

The term "CoRE kernel" stands for correlation-resemblance kernel. In many applications (e.g., vision), the data are often high-dimensional, sparse, and non-binary. We propose two types of (nonlinear) CoRE kernels for non-binary sparse data and demonstrate the effectiveness of the new kernels through a classification ex…

MKLpy simplifies Multiple Kernel Learning in Python.

Sparse Kernel Flows learns dynamical systems from data.

The study investigates kernel-target alignment in tree ensemble kernels.

Many real world graphs, such as the graphs of molecules, exhibit structure at multiple different scales, but most existing kernels between graphs are either purely local or purely global in character. In contrast, by building a hierarchy of nested subgraphs, the Multiscale Laplacian Graph kernels (MLG kernels) that we …

Two adaptive kernel selection methods improve the accuracy of Kernelized Diffusion Maps.

This study examines the practical equivalence of Laplace and neural tangent kernels.

The paper provides consistency results for KDE on manifolds with irregular kernels.