Origin identification of Cornus officinalis based on PCA-SVM combined model

Area of Science:

• Analytical Chemistry
• Chemometrics
• Spectroscopy

Background:

• Infrared spectroscopy offers rapid, non-destructive analysis for authenticity and origin identification.
• Chinese herbal medicines like Cornus officinalis require reliable origin verification for quality control.

Purpose of the Study:

• To establish an origin identification model for Cornus officinalis using mid-infrared spectroscopy.
• To evaluate the performance of a Principal Component Analysis-Support Vector Machine (PCA-SVM) model against other classification methods.

Main Methods:

• Mid-infrared spectral data from Cornus officinalis across 11 origins were collected.
• Principal Component Analysis (PCA) reduced spectral data dimensionality, retaining over 99.8% of information.
• A Support Vector Machine (SVM) classifier was trained using PCA-reduced data (PCA-SVM model).

Main Results:

• The PCA-SVM model achieved an accuracy of 84.8% for Cornus officinalis origin identification.
• PCA-SVM demonstrated superior performance in Accuracy, F1-Score, and Kappa coefficient compared to Naive Bayes, Decision Trees, LDA, RBFNN, and PLSDA.
• The PCA-SVM model showed higher accuracy and reduced variable redundancy compared to a full-spectrum SVM model.

Conclusions:

• The PCA-SVM model is an effective and accurate method for identifying the origin of Cornus officinalis.
• Mid-infrared spectroscopy combined with PCA and SVM offers a robust approach for authentication of herbal medicines.
• This methodology provides a valuable tool for ensuring the quality and traceability of traditional Chinese medicines.