Unified Native Spaces in Kernel Methods

Xavier Emery [email protected], Emilio Porcu [email protected], Moreno Bevilacqua [email protected] · 2025

Read Unified Native Spaces in Kernel Methods by Xavier Emery [email protected], Emilio Porcu [email protected], Moreno Bevilacqua [email protected] (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-01825.txt.

Description

Xavier Emery, Emilio Porcu, and Moreno Bevilacqua present a 2025 arXiv paper unifying kernel models used across statistics and machine learning. Many named kernel families exist, each tuned by parameters that control traits like smoothness, support range, and negative dips. The team builds one parametric class that contains the known families as special cases, exactly or as limits.

The authors also pin down the Sobolev space that matches the new kernel's own function space, and read off the spaces for older families as a by-product. The class can switch between compact and global support and can take negative values on real intervals. Famous members include the Matérn and Wendland kernels.

Catalog record

CREATORXavier Emery [email protected], Emilio Porcu [email protected], Moreno Bevilacqua [email protected]
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-01825

Browse the full library · About the Tower of Babel