Utilizing Explainable Artificial Intelligence (XAI) to Identify Determinants of Coffee Quality

This paper explores the transformative potential of Explainable Artificial Intelligence (XAI) in the context of coffee quality assessment, an area traditionally governed by subjective evaluation. By applying machine learning models, specifically a Random Forest Classifier enhanced by SHAP (SHapley Additive exPlanations) values, we identified crucial determinants of coffee quality, such as Category Two defects and high-altitude growth conditions. Our study demonstrates that machine learning can not only match but potentially exceed the accuracy of human experts in predicting coffee quality. More importantly, XAI has provided these models with a layer of transparency, making their complex predictions accessible and actionable for stakeholders in the coffee industry. This integration of AI into coffee quality assessment promises to standardize and optimize the evaluation process, offering a reliable guide for improving practices across the production chain. The findings underscore the broader impact of AI in agriculture, suggesting that such technology could be a harbinger of increased efficiency, sustainability, and trust in food production systems worldwide.