External parameter orthogonalization-support vector machine for processing of attenuated total...

Area of Science:

• Analytical Chemistry
• Chemometrics
• Food Science

Background:

• Saffron, the most expensive spice, faces authentication challenges due to adulteration.
• Existing methods for saffron authentication may lack sufficient accuracy and reliability.
• Attenuated Total Reflectance-Fourier Transform Mid-Infrared (ATR-FT-MIR) spectroscopy is a potential tool for analyzing saffron.

Purpose of the Study:

• To develop and validate a novel approach for authenticating saffron and detecting common adulterants.
• To process ATR-FT-MIR spectra effectively using external parameter orthogonalization (EPO) and support vector machine (SVM).

Main Methods:

• Principal Component Analysis (PCA) with EPO preprocessing was used for clustering authentic saffron samples.
• External Parameter Orthogonalization-Support Vector Machine (EPO-SVM) was developed for classifying binary mixtures of saffron and adulterants.
• Performance comparison with other classification methods was conducted.

Main Results:

• EPO-SVM achieved a classification accuracy greater than 95%, significantly outperforming other methods (<89.2%).
• Validation on mixed samples and commercial saffron demonstrated high classification figures of merit: sensitivity (96.6%), specificity (97.1%), and accuracy (96.8%).

Conclusions:

• The proposed EPO-SVM approach is a reliable and accurate tool for saffron authentication.
• EPO-PCA and EPO-SVM offer effective strategies for detecting adulteration in saffron samples.