EUVIMEDEUVIMED entdecken
Open Access · CC BY-ND 4.0

Probabilistic Nearest Neighbors Based Locality Preserving Projections for Unsupervised Metric Learning

Alaor Cervati Neto, Alexandre L. M. Levada

JUCS - Journal of Universal Computer Science · 28.05.2024 · 30 (5) · 603-616

DOI: 10.3897/jucs.107081 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JUCS - Journal of Universal Computer Science, online erschienen am 28.05.2024. Verfasst von Alaor Cervati Neto und 1 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-ND 4.0). Gefördert von: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Dimensionality reduction based unsupervised metric learning consists in finding meaningful compact data representations previously to clustering and classification problems. One of the major aspects of these algorithms is the approximation of the underlying manifold by a weighted graph. A limitation with most manifold learning algorithms is that edge weights in the proximity graph rely heavily on the Euclidean distance, which is known to be quite sensitive to the presence of outliers. In this paper, we propose to improve the Locality Preserving Projections (LPP) algorithm by incorporating a recently proposed graph inference method called Probabilistic Nearest Neighbors (PNN), an extension of the Clustering with Adaptive Neighbors (CAN) approach, used with success in graph-based semi-supervised learning. The proposed PNN-LPP algorithm is able to achieve better classification results than regular LPP, showing competitive performance against state-of-the-art approaches for dimensionality reduction, such as the UMAP algorithm, especially in datasets with a limited number of samples.

Förderung

  • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Zitieren

APA

Neto, A. C., & Levada, A. L. M. (2024). Probabilistic Nearest Neighbors Based Locality Preserving Projections for Unsupervised Metric Learning. JUCS - Journal of Universal Computer Science, 30(5), 603-616. https://doi.org/10.3897/jucs.107081

Vancouver

Neto AC, Levada ALM. Probabilistic Nearest Neighbors Based Locality Preserving Projections for Unsupervised Metric Learning. JUCS - Journal of Universal Computer Science. 2024;30(5):603-616. doi:10.3897/jucs.107081

BibTeX
@article{Neto2024,
  author = {Neto, Alaor Cervati and Levada, Alexandre L. M.},
  title = {Probabilistic Nearest Neighbors Based Locality Preserving Projections for Unsupervised Metric Learning},
  journal = {JUCS - Journal of Universal Computer Science},
  year = {2024},
  volume = {30},
  number = {5},
  pages = {603--616},
  doi = {10.3897/jucs.107081}
}
Kostenlos starten →