EUVIMEDEUVIMED entdecken
Crossref

NONPARAMETRIC REGRESSION FOR STRONGLY DEPENDENT RANDOM FIELDS

Djakaridja Diabate, S.-H. Arnaud Kanga, Ouagnina Hili, Sophie Dabo-Niang

Far East Journal of Theoretical Statistics · 2026 · 70 (4) · 635-658

DOI: 10.17654/0972086326030 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Far East Journal of Theoretical Statistics, 2026. Verfasst von Djakaridja Diabate und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper investigates the asymptotic behavior of two nonparametric estimators designed for spatial data that exhibit long range dependence. We work under a long memory assumption imposed jointly on the explanatory field and on the response variable. We study a conditional mean estimator, obtained as a spatial extension of the Nadaraya-Watson estimator, and a conditional quantile estimator; both incorporate spatial proximity explicitly through a location dependent weighting kernel, so that nearby sites contribute more to the prediction than distant ones. We establish almost sure consistency of the two estimators under strong spatial dependence, and illustrate these results through a simulation study.

Zitieren

APA

Diabate, D., Kanga, S. A., Hili, O., & Dabo-Niang, S. (2026). NONPARAMETRIC REGRESSION FOR STRONGLY DEPENDENT RANDOM FIELDS. Far East Journal of Theoretical Statistics, 70(4), 635-658. https://doi.org/10.17654/0972086326030

Vancouver

Diabate D, Kanga SA, Hili O, Dabo-Niang S. NONPARAMETRIC REGRESSION FOR STRONGLY DEPENDENT RANDOM FIELDS. Far East Journal of Theoretical Statistics. 2026;70(4):635-658. doi:10.17654/0972086326030

BibTeX
@article{Diabate2026,
  author = {Diabate, Djakaridja and Kanga, S.-H. Arnaud and Hili, Ouagnina and Dabo-Niang, Sophie},
  title = {NONPARAMETRIC REGRESSION FOR STRONGLY DEPENDENT RANDOM FIELDS},
  journal = {Far East Journal of Theoretical Statistics},
  year = {2026},
  volume = {70},
  number = {4},
  pages = {635--658},
  doi = {10.17654/0972086326030}
}
Kostenlos starten →