General Kernel Framework for Non-CND Distance Measures
A General Kernel Framework for Non-CND Distance Measures Using |D|-Dimensional Sparse Landmark Embeddings
This academic paper introduces a new theoretical framework for kernel methods that could improve the performance of Gaussian Processes on complex data types.
Researchers propose a new kernel method called the Sparse Landmark Embedding (SLE) kernel.
This framework allows using any distance measure, including geodesic and Wasserstein distances, in Gaussian Processes.
The method embeds inputs into a sparse feature space using training points, which ensures the kernel matrix is positive semi-definite.
Experiments show it matches or exceeds baselines in accuracy and uncertainty quantification.
A General Kernel Framework for Non-CND Distance Measures Using |D|-Dimensional Sparse Landmark Embeddings
Kernel methods, and Gaussian Processes (GPs) in particular, require a Hilbertian distance measure---one whose square is conditionally negative definite (CND)---to guarantee positive semi-definiteness (PSD) of the kernel matrix; a condition that fails for many natural input spaces, including smooth manifolds and spaces of probability distributions. We propose the Sparse Landmark Embedding (SLE) kernel, which eliminates this requirement entirely. Each input is embedded into a sparse feature vector via compactly supported bump functions centered at all |D| training points; applying any standard PSD kernel in this embedding space yields a kernel that is provably PSD for arbitrary distance measures. The compact support automatically controls embedding sparsity, keeping kernel matrices well-conditioned and computationally tractable despite the high ambient dimension. We provide theoretical guarantees on PSD, sparsity, stability, and universal approximation, and demonstrate, using geodesic and Wasserstein distances, that the SLE kernel matches or substantially exceeds domain-specific baselines in both predictive accuracy and uncertainty quantification.
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Kernel methods, Gaussian Processes, uncertainty quantification.
Source attribution: curated from arXiv cs.LG (@Marcus M. Noack, Maher B. Alghalayini, Mark D. Risser).