Preliminary diagnosis of pulmonary tuberculosis using ensemble method

摘要: Tuberculosis (TB) is the oldest human diseases with highest mortality rates among infectious disease. Indonesia fifth TB burden country in world. Diagnosis of difficult, especially case pediatric patients, extrapulmonary and smear-negative pulmonary TB. In addition, some tuberculosis symptoms have common not only lung cancer but also other diseases. This situation will lead to a delay correct diagnosis exposure inappropriate medication. Finally, individuals that receive inadequate treatment are more vulnerable develop multidrug-resistant tuberculosis. study aims model preliminary Preliminary established by using patient demographic data, anamnesis, physical examination. Some experiments were conducted classification techniques. single classifiers such as C4.5, Naive Bayes, Backpropagation SVM be compared ensemble method order improve performance model. The data taken from medical record patients Jakarta Respiratory Center. result showed provided best accuracy classifier.