Naïve Bayes classifier based on reliability measurement for datasets with noisy labels

Yingqiu Zhu (), Yinzhi Wang (), Lei Qin (), Bo Zhang (), Ben-Chang Shia () and MingChih Chen ()
Additional contact information
Yingqiu Zhu: University of International Business and Economics
Yinzhi Wang: University of International Business and Economics
Lei Qin: University of International Business and Economics
Bo Zhang: Renmin University of China
Ben-Chang Shia: Fu Jen Catholic University
MingChih Chen: Fu Jen Catholic University

Annals of Operations Research, 2025, vol. 349, issue 1, No 13, 259-286

Abstract: Abstract Incorrect labeling is a common issue that often occurs in machine learning applications. If datasets contain noisy labels and these errors are not corrected, the performance of the trained classifiers is affected significantly. In order to address this issue, we present a reliability measurement for labels, which is generated based on crowdsourcing. We adopt this reliability measurement to improve the Naïve Bayes classifier, resulting in a reliability measurement-based approach. Additionally, we explain the generating mechanism of incorrect labels and employ an iterative EM algorithm to optimize the corresponding log-likelihood function. This enables us to estimate the necessary parameters for the reliability measurement-based Naïve Bayes classifier. The simulation and experimental results demonstrate that the proposed method significantly improves the performance of Naïve Bayes classifier for datasets containing noisy labels.

Keywords: Naïve Bayes; Noisy label; Reliability measurement (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10479-023-05671-1 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:annopr:v:349:y:2025:i:1:d:10.1007_s10479-023-05671-1

Ordering information: This journal article can be ordered from
http://www.springer.com/journal/10479

DOI: 10.1007/s10479-023-05671-1

Access Statistics for this article

Annals of Operations Research is currently edited by Endre Boros

More articles in Annals of Operations Research from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().