Detecting carob powder adulteration in cocoa using near and mid-infrared spectroscopy: A comprehensive classification and regression analysis

Basit Öğe Kaydı
dc.authoridAyvaz, Hüseyin / 0000-0001-9705-6921
dc.authoridDoğan, Muhammed Ali / 0000-0002-5524-7567
dc.contributor.authorTurgut, Sebahattin Serhat
dc.contributor.authorAyvaz, Hüseyin
dc.contributor.authorDoğan, Muhammed Ali
dc.contributor.authorMarin, Dolores Perez
dc.contributor.authorMenevşeoğlu, Ahmed
dc.date.accessioned2025-05-29T02:58:00Z
dc.date.available2025-05-29T02:58:00Z
dc.date.issued2025
dc.departmentÇanakkale Onsekiz Mart Üniversitesi
dc.description.abstractCocoa powder is a globally traded food product, primarily produced in developing nations, with substantial economic importance. However, it is susceptible to adulteration with inexpensive materials such as carob flour, particularly in low concentrations too low to be detected by sensory methods. To address this issue, rapid analytical techniques such as vibrational spectroscopy combined with multivariate analysis could be beneficial for rapid and reliable detection of adulteration. In this study, spectral data were collected using four different infrared spectrometers: a benchtop FT-NIR system, two portable NIR instruments, and a benchtop FT-MIR-ATR. Samples included pure cocoa, pure carob, and their mixtures with carob concentrations ranging from 0 % to 60 %. Both classification and regression models were developed to detect and quantify the presence of carobs in cocoa powder. Classification models, including Random Forest (RF), Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), k-Nearest Neighbors (kNN), Linear Discriminant Analysis (LDA), and Soft Voting Classifiers, demonstrated superior performance for discriminating between cocoa powder, carob powder, and cocoa-carob mixtures with the area under the receiver operating characteristic curve (AUC) scores achieving the level of higher than 0.99, particularly using the benchtop FT-NIR, one of the cost-effective portable NIR and FT-MIR-ATR devices. Similarly, regression models - RF, SVM, MLP, kNN, Partial Least Squares Regression, and Voting Regressor- exhibited robust predictive capabilities. Particularly, FT-MIR and portable NIR based models showed exceptional accuracy with RPD values exceeding 16 and 13, respectively, signifying their applicability in quality and process control. Key wavelength driving model predictions were identified using permutation feature importance for both regressors and classifiers. Overall, these findings highlight and prove the potential of NIR and MIR spectroscopy as rapid, robust, and non-destructive tools for screening and quality control in food authentication.
dc.description.sponsorshipCOST Action, SensorFINT [CA19145]; COST Action [CA19145]
dc.description.sponsorshipAhmed Menevseoglu received a STSM grant from COST Action-CA19145, SensorFINT, and this study was partially supported by COST Action-CA19145.
dc.identifier.doi10.1016/j.foodres.2025.116132
dc.identifier.issn0963-9969
dc.identifier.issn1873-7145
dc.identifier.pmid40263821
dc.identifier.scopus2-s2.0-86000651801
dc.identifier.scopusqualityQ1
dc.identifier.urihttps://doi.org/10.1016/j.foodres.2025.116132
dc.identifier.urihttps://hdl.handle.net/20.500.12428/30236
dc.identifier.volume208
dc.identifier.wosWOS:001446432200001
dc.identifier.wosqualityQ1
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakScopus
dc.indekslendigikaynakPubMed
dc.language.isoen
dc.publisherElsevier
dc.relation.ispartofFood Research International
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.snmzKA_WOS_20250529
dc.subjectCocoa
dc.subjectCarob
dc.subjectNon-destructive analysis
dc.subjectNIR
dc.subjectMIR
dc.subjectChemometrics
dc.titleDetecting carob powder adulteration in cocoa using near and mid-infrared spectroscopy: A comprehensive classification and regression analysis
dc.typeArticle

Dosyalar

Orijinal paket
Listeleniyor 1 - 1 / 1
[ X ]
İsim:
Huseyin Ayvaz_Makale.pdf
Boyut:
7.28 MB
Biçim:
Adobe Portable Document Format
İndir

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
Gıda Mühendisliği Bölümü Koleksiyonu
PubMed İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu