EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Weighted Chernoff Information and Optimal Loss Exponent in Context-Sensitive Hypothesis Testing

Mark Kelbert iD, El'mira Yu Kalimulina iD

Entropy · 08.05.2026 · 28 (5) · 536

DOI: 10.3390/e28050536 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Entropy, online erschienen am 08.05.2026. Verfasst von Mark Kelbert und 1 weiteren Autor:innen. Beteiligte Einrichtung: Laboratory of Stochastic Analysis and Its Applications, Department of Statistics and Data Analysis, National Research University Higher School of Eco…. Die Arbeit nennt 17 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: HSE University, Ministry of Science and Higher Education of the Russian Federation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We study binary hypothesis testing for i.i.d. observations under a multiplicative context weight. For the optimal weighted total loss, defined as the sum of weighted type-I and type-II losses, we prove the logarithmic asymptotic Ln∗=exp{−nDCw(P,Q)+o(n)},n→∞, where DCw is the weighted Chernoff information. The single-letter form of the exponent relies on a structural assumption that the weight factorises across observations, φ(x1n)=∏i=1nφ(xi); this restriction is essential for the single-letter representation and should be distinguished from the weaker qualitative description “multiplicative context weight”. The proof embeds the weighted geometric mixtures φpαq1−α into a likelihood-ratio exponential family and identifies the rate through its log-normaliser. We also derive concentration bounds for the tilted weighted log-likelihood, obtain closed forms for Gaussian, Poisson, and exponential models, and extend the exponent characterisation to finitely many hypotheses.

Schlagwörter

Stichwörter der Autor:innen

context-sensitive lossexponential familyhypothesis testinginformation geometryweighted Bhattacharyya coefficientweighted Chernoff information

Autor:innen und Institutionen

  • Laboratory of Stochastic Analysis and Its Applications, Department of Statistics and Data Analysis, National Research University Higher School of Economics, 101000 Moscow, Russia.
  • Department of Mathematics, Swansea University, Swansea SA2 8PP, UK.
  • Institute for Information Transmission Problems, Russian Academy of Sciences (IITP RAS), 127051 Moscow, Russia.
  • Faculty of Mechanics and Mathematics, Lomonosov Moscow State University, 119991 Moscow, Russia.

ORCID-Profile: Mark Kelbert, El'mira Yu Kalimulina

Förderung

  • HSE University
  • Ministry of Science and Higher Education of the Russian Federation

Zitieren

APA

Kelbert, M., & Kalimulina, E. Y. (2026). Weighted Chernoff Information and Optimal Loss Exponent in Context-Sensitive Hypothesis Testing. Entropy, 28(5), 536. https://doi.org/10.3390/e28050536

Vancouver

Kelbert M, Kalimulina EY. Weighted Chernoff Information and Optimal Loss Exponent in Context-Sensitive Hypothesis Testing. Entropy. 2026;28(5):536. doi:10.3390/e28050536

BibTeX
@article{Kelbert2026,
  author = {Kelbert, Mark and Kalimulina, El'mira Yu},
  title = {Weighted Chernoff Information and Optimal Loss Exponent in Context-Sensitive Hypothesis Testing},
  journal = {Entropy},
  year = {2026},
  volume = {28},
  number = {5},
  pages = {536},
  doi = {10.3390/e28050536}
}
Kostenlos starten →