Geographical Discriminant and Classification of Cinnamomum Cassia Collected in Vietnam Using Atr-ftir Coupled With Machine Learning Algorithms
Year: 2025
Type: article
Abstract: In Vietnam, Cinnamomum cassia is widely utilized not only as a spice but also as a key ingredient in medicine, cosmetics, and the processing industry. The varying climatic conditions across different geographic regions where cinnamon is cultivated significantly influence its quality, often leading to the unintentional mixing of cinnamon from multiple sources. To analyze cinnamon samples, Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR) was employed to measure 139 samples collected from four distinct provinces in Vietnam – Yen Bai, Quang Ninh, Thanh Hoa, and Quang Nam – each representing unique climatic conditions. To enhance data quality, the second derivative Savitzky-Golay smoothing method was applied, improving spectral resolution and reducing noise. Unsupervised chemometric techniques, including Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA), were used to assess classification potential. Additionally, supervised machine learning models, such as Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), GausianNB, BernoulliNB, AdaBoost, and Gradient Boosting were combined with PCA results and trained using first derivative Savitzky-Golay smoothed spectra for classification. The results demonstrated that the PCA-SVM model achieved the highest classification accuracy (96.88%) with the second derivative spectra, whereas first derivative and raw data exhibited lower performance. This study underscores that ATR-FTIR spectroscopy, when combined with second-derivative preprocessing and PCA-SVM, provides a simple, rapid, non-destructive, and cost-effective approach for distinguishing cinnamon bark samples based on their geographical origin, offering valuable insights for quality assessment. For citation: Bui Thi Lan Phuong, Hoang Thi Bich, Nguyen Van Phuong, Bui An Duy, Nguyen Duc Phong, Nguyen Manh Son, Pham Gia Bach, Nguyen Thi Cam Ha, Ta Thi Thao, Nguyen Thi Kieu Anh Geographical discriminant and classification of cinnamomum cassia collected in Vietnam using ATR-FTIR coupled with machine learning algorithms. ChemChemTech [Izv. Vyssh. Uchebn. Zaved. Khim. Khim. Tekhnol.]. 2025. V. 68. N 7. P. 102-113. DOI: 10.6060/ivkkt.20256807.7246.
Authors: Bui Thi Lan Phuong Phuong, Hoang Thi Bich, Nguyen Van Phuong, Bui An Duy, Nguyen Duc Phong, Nguyen Manh Son, Pham Gia Bach, Nguyen Thi Cam Ha, Ta Thi Thao, Nguyen Thi Kieu Anh
Institutions: Hanoi University of Pharmacy, VNU University of Science
Language: English
Subfield: Analytical Chemistry
Field: Chemistry
Domain: Physical Sciences
SDG: Zero hunger
Open Access status: diamond
OpenAlex ID: W4410899763