Tuberkulosis merupakan penyakit menular yang masih menjadi permasalahan kesehatan di Indonesia sehingga diperlukan metode yang lebih cepat dan akurat untuk membantu proses diagnosis. Penelitian ini bertujuan untuk membandingkan kinerja algoritma Support Vector Machine (SVM) dan Random Forest pada klasifikasi penyakit TBC menggunakan data primer sebanyak 201 pasien di Puskesmas Sandar Angin Pagar Alam periode Januari 2023 hingga Oktober 2025. Penelitian menggunakan 11 variabel, yaitu nyeri dada, batuk berdahak, lama batuk, sesak napas, kelelahan, penurunan berat badan (kg), suhu tubuh (°C), keringat dingin, nafsu makan berkurang, sputum, dan diagnosis. Evaluasi model dilakukan menggunakan pembagian data 80:20 dan 5-Fold Cross Validation. Pada pembagian data 80:20, SVM memperoleh accuracy 98%, precision 94%, recall 100%, dan F1-score 97%, sedangkan Random Forest memperoleh accuracy 93%, precision 88%, recall 94%, dan F1-score 91%. Pada 5-Fold Cross Validation, SVM juga menunjukkan accuracy terbaik sebesar 90%, lebih tinggi dibandingkan Random Forest sebesar 88%. Hasil penelitian menunjukkan bahwa algoritma SVM memiliki performa klasifikasi yang lebih baik dan lebih konsisten dibandingkan Random Forest, sehingga lebih direkomendasikan untuk klasifikasi penyakit TBC pada data pasien di Puskesmas Sandar Angin Pagar Alam.

Tuberculosis remains one of the major public health problems in Indonesia, highlighting the need for faster and more accurate approaches to support the diagnostic process. This study aims to compare the performance of the Support Vector Machine (SVM) and Random Forest algorithms in classifying tuberculosis using primary data collected from 201 patients at Sandar Angin Public Health Center, Pagar Alam, from January 2023 to October 2025. The study employed 11 variables, namely chest pain, productive cough, cough duration, shortness of breath, fatigue, weight loss (kg), body temperature (°C), night sweats, loss of appetite, sputum, and diagnosis. Model evaluation was conducted using an 80:20 train-test split and 5-Fold Cross Validation. Using the 80:20 split, the SVM algorithm achieved an accuracy of 98%, precision of 94%, recall of 100%, and an F1-score of 97%, while the Random Forest algorithm achieved an accuracy of 93%, precision of 88%, recall of 94%, and an F1-score of 91%. In the 5-Fold Cross Validation evaluation, SVM also demonstrated the highest accuracy of 90%, outperforming Random Forest, which achieved 88%. The results indicate that the SVM algorithm provides better and more consistent identification performance than Random Forest, making it the more suitable algorithm for tuberculosis identification using patient data from Sandar Angin Public Health Center, Pagar Alam.