Tuberculosis (TB) is an infectious disease cause by Mycobacterium tuberculosis that typically affects the lungs, but also can affect any organ in the body. This study aims to obtain a screening model for early detection of tuberculosis using the multi-objective gradient evolution-based support vector machine and c5.0 decision tree method based on tuberculosis risk factors on Palembang City. This was a case-control study with an analytic observational design. Data were collected by interview and physical examination on 240 respondents for each case and control group. Analysis for the risk factors of tuberculosis in this study used binary logistic regression analysis. Establishment of TB early detection model using gradient evolution algorithm. In this study, the majority of respondents with TB were male (60%), 35 - 39 years old (70%), had elementary school education (60%), were underweight (78.6%), had sufficient income (54,7%), were active smokers (68.6%), lived in the same house as TB patients (94.9%), lived in a densely populated house (54.6%), and did not have BCG immunization (59.5%). Risk factors for TB incidence in Palembang City were statistically significant for gender, education, nutritional status, smoking habits, household contact, family size, and BCG immunization while age were not risk factors for TB in Palembang City. Based on the TB screening model using the GE-SVM algorithm, the accuracy rate for training data is 80.83%, while it is 74.61% for testing data. The simulation results show that the model produced in this study can help medical personnel in conducting initial screening of TB risk to a person.