EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Non-parametric density estimation based on Kernel, Wavelets, and K-Means

Raquel Romes Linhares, Regis Nunes Vargas

Revista DCS · 16.07.2026 · 23 (92) · e6014

DOI: 10.54899/dcs.v23i92.6014 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Revista DCS, online erschienen am 16.07.2026. Verfasst von Raquel Romes Linhares und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This work presents a methodology for non-parametric density estimation based on the integration of the Gaussian kernel estimator, wavelet multiresolution analysis, and clustering techniques applied to wavelet coefficients. Traditional methods often struggle to effectively separate noise from relevant density structures without relying on subjective thresholding choices. To overcome this challenge, the proposed approach initially obtains a smooth estimate using the kernel estimator. Next, this preliminary estimate is decomposed into approximation and detail wavelet coefficients. Instead of using standard thresholds, the detail coefficients are clustered using the k-means algorithm. This data-driven strategy allows for the automatic identification and separation of components associated with noise from those representing relevant underlying structures. Finally, the density function is reconstructed using only the informative coefficients. The results confirm that this methodology acts as an adaptive filtering mechanism, enabling the reconstruction of functions that preserve critical local characteristics while simultaneously reducing noise. An illustrative example was included to demonstrate the effectiveness of the method.

Zitieren

APA

Linhares, R. R., & Vargas, R. N. (2026). Non-parametric density estimation based on Kernel, Wavelets, and K-Means. Revista DCS, 23(92), e6014. https://doi.org/10.54899/dcs.v23i92.6014

Vancouver

Linhares RR, Vargas RN. Non-parametric density estimation based on Kernel, Wavelets, and K-Means. Revista DCS. 2026;23(92):e6014. doi:10.54899/dcs.v23i92.6014

BibTeX
@article{Linhares2026,
  author = {Linhares, Raquel Romes and Vargas, Regis Nunes},
  title = {Non-parametric density estimation based on Kernel, Wavelets, and K-Means},
  journal = {Revista DCS},
  year = {2026},
  volume = {23},
  number = {92},
  pages = {e6014},
  doi = {10.54899/dcs.v23i92.6014}
}
Kostenlos starten →