Detecting Honey Adulteration: Advanced Approach Using UF-GC Coupled with Machine Learning

Area of Science:

• Food Chemistry
• Analytical Chemistry
• Data Science

Background:

• Honey adulteration poses a significant threat to consumer trust and market integrity.
• Existing detection methods can be time-consuming, costly, or lack comprehensive accuracy.
• The need for rapid, reliable, and cost-effective analytical techniques for food authenticity is paramount.

Purpose of the Study:

• To develop and validate a novel method for detecting honey adulteration.
• To integrate ultra-fast gas chromatography (UF-GC) with machine learning (ML) for enhanced analytical performance.
• To assess the efficacy of Support Vector Regression (SVR) and Least Absolute Shrinkage and Selection Operator (LASSO) models in predicting honey adulteration.

Main Methods:

• Ultra-fast gas chromatography (UF-GC) was employed for rapid sample analysis.
• Machine learning models, specifically SVR and LASSO, were trained and applied to GC data.
• Models were evaluated for their predictive accuracy in detecting adulteration in orange blossom (OB) and sunflower (SF) honeys.

Main Results:

• The SVR model demonstrated high predictive power (R² > 0.90) for combined honey types.
• Treating OB and SF honeys individually significantly improved accuracy, with R² values exceeding 0.99.
• The LASSO model showed particular effectiveness when applied to individual honey types.

Conclusions:

• The integration of UF-GC and ML provides a robust and reliable method for honey adulteration detection.
• This approach offers a significant advancement over traditional methods, ensuring honey authenticity.
• The methodology holds potential for application in authenticating other food products, advancing food safety standards.