Regression Modelling with Coupled GC–MS/GC–FID for Robust Classification of Aquilaria Species

Abstract

The Aquilaria genus is renowned for producing agarwood, a highly valuable resinous heartwood with diverse applications in perfumery, traditional medicine, and cultural practices. Accurate species identification is critical for quality control, consumer trust, and conservation, yet traditional methods are often unreliable due to chemical similarities among species. This study aimed to develop a regression-based classification model for four Aquilaria species (A. beccariana, A. malaccensis, A. crassna, and A. subintegra) using the chemical composition of their essential oils. Essential oils were extracted via hydrodistillation and analyzed using Gas Chromatography–Mass Spectrometry (GC–MS) coupled with Gas Chromatography–Flame Ionization Detection (GC–FID). Stepwise multiple regression identified three chemical compounds as the most significant contributors to species differentiation. The final model achieved a high predictive accuracy (R² = 0.990; adjusted R² = 0.989), with clear visual separation of species observed in three-dimensional scatter plots. These findings demonstrate that a targeted multivariate approach using a small set of chemical markers provides a robust, interpretable, and scalable method for Aquilaria species classification, offering valuable applications for authentication, quality control, and conservation in the agarwood industry.