Hypothesis Testing in Nonparametric Regression Models Using Fourier Series for Trigonometric Polynomial Representation
Boletim da Sociedade Paranaense de Matemática · 02.07.2026 · 44 (7) · 1-15
DOI: 10.5269/bspm.82660 ↗Kurz gefasst
Fachartikel in der Zeitschrift Boletim da Sociedade Paranaense de Matemática, online erschienen am 02.07.2026. Verfasst von Rakia Ahmed Yahia und 1 weiteren Autor:innen. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This paper addresses the problem of hypothesis testing in nonparametric regression models using Fourier series techniques. The central objective is to test whether the unknown regression function can be adequately represented as a trigonometric polynomial of a given degree. To this end, we employ a Fourier series decomposition to approximate the regression function and estimate its Fourier coefficients from the observed data. The proposed methodology begins with the estimation of the regression function by truncating its Fourier expansion to a finite number of terms, reflecting the assumption of smoothness and periodicity in the underlying function. The estimated coefficients serve as the basis for constructing test statistics aimed at verifying whether the function lies within the class of trigonometric polynomials of degree at most s. The null hypothesis posits that higher-order Fourier coefficients beyond degree s are statistically insignificant. We develop a testing procedure grounded in the asymptotic properties of the Fourier coefficients under the null and alternative hypotheses. The test’s performance is evaluated through extensive simulations under various scenarios, including different sample sizes, noise levels, and function complexities. Results indicate that the proposed Fourier-based test exhibits strong discriminatory power and robustness in identifying deviations from the hypothesized polynomial structure. This study highlights the effectiveness of Fourier methods in model specification testing within the nonparametric regression framework and provides a foundation for future extensions to multivariate or non-periodic contexts.
Zitieren
APA
Yahia, R. A., & Zeghdoudi, H. (2026). Hypothesis Testing in Nonparametric Regression Models Using Fourier Series for Trigonometric Polynomial Representation. Boletim da Sociedade Paranaense de Matemática, 44(7), 1-15. https://doi.org/10.5269/bspm.82660
Vancouver
Yahia RA, Zeghdoudi H. Hypothesis Testing in Nonparametric Regression Models Using Fourier Series for Trigonometric Polynomial Representation. Boletim da Sociedade Paranaense de Matemática. 2026;44(7):1-15. doi:10.5269/bspm.82660
BibTeX
@article{Yahia2026,
author = {Yahia, Rakia Ahmed and Zeghdoudi, Halim},
title = {Hypothesis Testing in Nonparametric Regression Models Using Fourier Series for Trigonometric Polynomial Representation},
journal = {Boletim da Sociedade Paranaense de Matemática},
year = {2026},
volume = {44},
number = {7},
pages = {1--15},
doi = {10.5269/bspm.82660}
}
EUVIMED