EUVIMEDEUVIMED entdecken
Crossref

A general framework for divergence approximation on Gaussian mixture models

Amit Vishwakarma, K. S. Subrahamanian Moosath

Information Geometry · 2026-11 · 9 (S1) · 51-77

DOI: 10.1007/s41884-026-00203-9 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Information Geometry, online erschienen am 07.05.2026. Verfasst von Amit Vishwakarma und 1 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Widely used divergences like Kullback–Leibler (KL) divergence, Bhattacharyya divergence and Cauchy–Schwarz divergence have no closed form expression in the case of Gaussian mixture models (GMMs). This led to computationally expensive methods like numerical approximations, componentwise methods which fails to capture the mixture structure. In this paper, we develop a divergence approximation through the embedding of GMMs into the manifold of symmetric positive definite (SPD) matrices. The main result is that for regular divergences on compact set of non-degenerate GMM parameters, uniform equivalence is obtained for the divergence between two GMMs and computationally tractable corresponding divergence between their centered multivariate normal distribution representations in the SPD manifold. This is an extension of the work presented in GSI’25 conference. We further prove a stability theorem showing that the uniform equivalence degrades only linearly under spectral perturbations of the Hessian, ensuring robustness in numerical implementation. Experiments on the UIUC and KTH-TIPS texture recognition benchmarks evaluate three divergence measures computed via the SPD embedding: symmetric KL divergence, Bhattacharyya divergence and Cauchy–Schwarz divergence.

Zitieren

APA

Vishwakarma, A., & Moosath, K. S. S. (2026). A general framework for divergence approximation on Gaussian mixture models. Information Geometry, 9(S1), 51-77. https://doi.org/10.1007/s41884-026-00203-9

Vancouver

Vishwakarma A, Moosath KSS. A general framework for divergence approximation on Gaussian mixture models. Information Geometry. 2026;9(S1):51-77. doi:10.1007/s41884-026-00203-9

BibTeX
@article{Vishwakarma2026,
  author = {Vishwakarma, Amit and Moosath, K. S. Subrahamanian},
  title = {A general framework for divergence approximation on Gaussian mixture models},
  journal = {Information Geometry},
  year = {2026},
  volume = {9},
  number = {S1},
  pages = {51--77},
  doi = {10.1007/s41884-026-00203-9}
}
Kostenlos starten →