Date of Publication
8-2025
Document Type
Bachelor's Thesis
Degree Name
Bachelor of Science in Chemistry
Subject Categories
Chemistry
College
College of Science
Department/Unit
Chemistry
Thesis Advisor
Emmanuel V. Garcia
Defense Panel Chair
Lourdes P. Guidote
Defense Panel Member
Mariafe N. Calingacion
Vladimir U. William
Abstract (English)
Coffee is a standard beverage often consumed to boost mental alertness due to its caffeine content. In the Philippines, coffee production has declined significantly in the past years due to urbanization and land conversion. On the contrary, the demand for coffee is expected to increase significantly in the following years. A low production yield gives rise to fraudulent activities related to mislabeling coffee beans' botanical and geographical origins. Thus, this paper aims to classify coffee beans based on their origin and variety through multi-elemental analysis. Thirty-eight (38) samples were gathered all over Mindanao and were subjected to oven drying for 24 hours at 60℃. These samples were then pulverized and pelletized before being analyzed using a portable ED-XRF (pXRF). There were 13 elements detected from analysis: K, Mg, P, S, Cl, Mn, Cr, Rb, Sr, Pd, Cu, Ni, and Zn. The findings reveal that Mn and Sr as elemental predictors consistently showed great importance for both varietal and origin. Nevertheless, clustering patterns for varietal tracing illustrated better separation as compared to geographical tracing. Difference in compositions for the species contributed to the clustering patterns observed. Random Forest (RF), a supervised machine learning technique, combined with LDA analysis proved to be an excellent discriminatory method for coffee beans in terms of their species and origin due to high classification accuracies and strong class metrics for all models.
Abstract Format
html
Abstract (Filipino)
"-"
Abstract Format
html
Language
English
Format
Electronic
Keywords
Coffee--Philippines; Coffee—Varieties; Machine learning
Recommended Citation
Salcor, C. B., & Tiu, M. M. (2025). Varietal and geographical tracing of Philippine coffee beans using multi-elemental analysis and machine learning techniques. Retrieved from https://animorepository.dlsu.edu.ph/etdb_chem/69
Upload Full Text
wf_yes
Embargo Period
8-8-2026
