Classification of the Relationship Between Fruit Consumption and Diabetes Risk Using the Naïve Bayes Algorithm

Abstract

Diabetes is a non-communicable disease whose prevalence continues to rise, where dietary patterns, including fruit intake, contribute to diabetes risk. This study classifies the relationship between fruit consumption and diabetes risk using the Naïve Bayes algorithm. The dataset consists of 400 synthetic records with 8 attributes. Data preprocessing included cleaning, normalization, and 5-Fold Cross Validation. Results show accuracy of 96.25%, precision 100%, recall 57%, and F1-score 0.73. The model was implemented into GlucoSense, a web-based system providing real-time diabetes risk predictions. This research proves that Naïve Bayes is effective for classifying diabetes risk based on fruit consumption patterns.