Swamping and masking rate in binomial regression model
Boletim da Sociedade Paranaense de Matemática · 27.03.2026
DOI: 10.5269/bspm.80163 ↗Kurz gefasst
Fachartikel in der Zeitschrift Boletim da Sociedade Paranaense de Matemática, online erschienen am 27.03.2026. Verfasst von Noora N. Saleh. Die Arbeit nennt 9 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
High-dimensional data detection becomes challenging due to swamping and masking effects since they prevent the identification of real outliers. This study demonstrates how the integration of robust statistical methods enhances the detection of fraudulent transactions using binomial regression and Isolation Forest models. Binomial regression, leveraging robust variable selection techniques, ensures optimal feature selection while mitigating the effects of swamping and masking. Isolation Forest, with carefully tuned subsampling, enhances anomaly detection by isolating fraudulent instances effectively. Performance evaluation is conducted using accuracy, precision, recall, F1-score, and AUC-ROC metrics, compared against traditional outlier detection methods such as K-Means and DBSCAN.
Zitieren
APA
Saleh, N. N. (2026). Swamping and masking rate in binomial regression model. Boletim da Sociedade Paranaense de Matemática. https://doi.org/10.5269/bspm.80163
Vancouver
Saleh NN. Swamping and masking rate in binomial regression model. Boletim da Sociedade Paranaense de Matemática. 2026. doi:10.5269/bspm.80163
BibTeX
@article{Saleh2026,
author = {Saleh, Noora N.},
title = {Swamping and masking rate in binomial regression model},
journal = {Boletim da Sociedade Paranaense de Matemática},
year = {2026},
doi = {10.5269/bspm.80163}
}
EUVIMED