Classification Analysis of QRIS Usage Generation Z With Naïve Bayes Algorithm

This paper examines factors related to the use of the Indonesian Standard Quick Response Code (QRIS) among Generation Z and builds a probabilistic classification model to distinguish between active and passive users.A total of 152 valid responses were collected from students through an online questionnaire using a Likert scale.After data cleaning and feature coding, the Naïve Bayes algorithm was trained using stratified training and testing data (75/25).The model achieved an accuracy of 47.22%, precision (weighted) of 51.91%, recall (weighted) of 47.22%, and F1-score (weighted) of 48.70%.Although its performance was moderate, the analysis revealed key determinants such as perceived ease, transaction speed, and perceived security.This study also discusses the causes of misclassification (class imbalance, mixed feature scales, and multiple choice variables) and provides suggestions for improvement (feature engineering, data balancing, and model comparison) for further research.