PREDICTION OF TUBERCULOSIS PATIENTS WITH MACHINE LEARNING ALGORITHMS

This research is highly significant because Tuberculosis remains a significant global health issue, and early detection can aid in its more effective management. By employing four different classification algorithms, this study provides a deep understanding of how each algorithm can contribute to Tuberculosis detection. The evaluation of four classification algorithms, namely Logistic Regression (LR), K-Nearest Neighbor (K-NN), Random Forest (RF), and Naive Bayes (NB), in detecting Tuberculosis (TB) was conducted using a dataset comprising various clinical and biological features related to Tuberculosis. The research findings indicate that the Random Forest and K-NN algorithms achieved the highest accuracy of 99.8%, followed by Logistic Regression with 99% accuracy and Naive Bayes. Considering these research findings, the next steps may involve the development of more efficient detection methods based on the combination or enhancement of the evaluated algorithms. Additionally, this research can also serve as a foundation for guiding efforts in early treatment planning for individuals infected with Tuberculosis