# Predicting Bioactive Compounds in Arbutus unedo L. Leaves Using Machine Learning: Influence of Extraction Technique, Solvent Type, and Geographical Location

## TL;DR

This study uses machine learning to predict bioactive compounds in Arbutus unedo leaves, finding that solvent type and extraction method significantly affect compound recovery.

## Contribution

The novel use of machine learning models to predict bioactive compound recovery based on extraction parameters and geographical origin.

## Key findings

- 70% ethanol yielded the highest levels of bioactives and antioxidant capacity.
- Leaves from Vis had higher total phenolics and condensed tannins than those from Mali Lošinj.
- Gradient Boosting and Decision Tree models achieved high accuracy (R2 > 0.91) in predicting bioactive compounds.

## Abstract

This study investigates the effects of extraction technique, solvent type, and geographical origin on the recovery of bioactive compounds from Arbutus unedo L. leaves collected from two Croatian islands (Vis and Mali Lošinj) and extracted using conventional, Soxhlet, and ultrasound-assisted extraction (UAE) with green solvents (distilled water, 70% ethanol, and ethyl acetate). Extracts were purified and characterized by thin-layer chromatography, column chromatography, and FTIR spectroscopy. Total phenols, hydroxycinnamic acids, flavonols, condensed tannins, and antioxidant capacity were quantified spectrophotometrically. Solvent type had the greatest influence, with 70% ethanol yielding the highest levels of bioactives and antioxidant capacity. Geographical origin significantly affected total phenolics and condensed tannins, with leaves from Vis outperforming those from Mali Lošinj. UAE was slightly more efficient than conventional and Soxhlet methods, particularly for thermolabile phenolics. Machine learning algorithms were applied as exploratory tools, using total phenols as a proxy variable to estimate selected bioactive compounds and antioxidant capacity based on extraction parameters. Decision Tree and Gradient Boosting models showed high goodness of fit within the experimental dataset (R2 > 0.91). These results support the potential of green extraction strategies combined with data-driven screening for the valorization of A. unedo leaf extracts, while highlighting the need for further validation prior to industrial application.