Comparison of Decision Tree and Naïve Bayes in Data Mining for Classifying Tuberculosis Disease Data in Indonesia

Tuberculosis (TB) remains a serious health problem in Indonesia, with high case and death rates. Indonesia has the second-highest number of TB cases in the world after India, accounting for approximately 10% of global cases. Data from the Indonesian Ministry of Health indicate that the number of TB cases is estimated to exceed one million per year, with hundreds of thousands of deaths. Therefore, better data analysis is needed to support TB control and prevention strategies. This study aims to classify the spread of TB in various districts/cities in five provinces with the highest TB cases in Indonesia by comparing the Decision Tree and Naïve Bayes methods using TB data from 2021 to 2024. The dataset includes variables such as age, population density, and health services. The research stages include data pre-processing, converting numerical data to categorical data, and validation using K-Fold Cross Validation. Model performance was evaluated using a Confusion Matrix. The results showed that the Decision Tree method had a higher accuracy rate (76.63%) than the Naïve Bayes method (63.27%). The superiority of Decision Tree in identifying age as the main predictor shows its ability to model complex TB data patterns.