Scholarlink

<p>This study proposes a fast and non-destructive method for the authentication of commercial honey samples using mid-infrared (MIR) spectroscopy combined with one-class classification algorithms. A total of 87 honey samples (19 unadulterated and 68 adulterated), seized by the Brazilian Federal Police, were analyzed. These samples were previously authenticated by the isotope ratio mass spectrometry (IRMS) reference method. An exploratory analysis by principal component analysis (PCA) revealed clear spectral differences between unadulterated and adulterated samples, highlighting the fingerprint region (1500–750&#xa0;cm<sup>−1</sup>) as the most informative spectral interval for authenticity assessment. One-class classification models were developed using Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) and One-Class Partial Least Squares (OC-PLS). The DD-SIMCA and OC-PLS models demonstrated excellent performance, achieving 100% sensitivity in the training set and 100% sensitivity and specificity in the test set. These results highlight the potential of MIR combined with DD-SIMCA and OC-PLS as a robust and reliable approach for assessing honey authenticity.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Authentication of Commercial Honey Samples Using Infrared Spectroscopy Combined with One-Class Classifiers

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月1日

查看全文
  • Ricardo de Oliveira Mascarenhas,
  • Maria Eduarda Negrello,
  • Marcus Vinícius de Oliveira Andrade,
  • Poliana M. Santos

摘要

This study proposes a fast and non-destructive method for the authentication of commercial honey samples using mid-infrared (MIR) spectroscopy combined with one-class classification algorithms. A total of 87 honey samples (19 unadulterated and 68 adulterated), seized by the Brazilian Federal Police, were analyzed. These samples were previously authenticated by the isotope ratio mass spectrometry (IRMS) reference method. An exploratory analysis by principal component analysis (PCA) revealed clear spectral differences between unadulterated and adulterated samples, highlighting the fingerprint region (1500–750 cm−1) as the most informative spectral interval for authenticity assessment. One-class classification models were developed using Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) and One-Class Partial Least Squares (OC-PLS). The DD-SIMCA and OC-PLS models demonstrated excellent performance, achieving 100% sensitivity in the training set and 100% sensitivity and specificity in the test set. These results highlight the potential of MIR combined with DD-SIMCA and OC-PLS as a robust and reliable approach for assessing honey authenticity.