Background: Health is one of the vital factors for human survival and continuous efforts are focused on research in this domain. Spreading of tuberculosis increases from year to year and is a life threat that exists since antiquity. Problem: The problem gets complicated with other viruses like HIV and drug-resistant tuberculosis. World Health Organization has published data about drug-resistant tuberculosis to analyze the problems faced. Objective: This paper focuses on applying a semi-supervised learning model to prescribe recommendations based on data analytics. Three models such as the Decision Tree, Gradient Boosting and Neural Network are trained to predict the clusters. Gradient Boosting can perform the best with the lowest misclassification rate and the majority cluster is identified based on its impact and population. Conclusion: The outcome of this analysis can provide recommendations to the health domain to reduce the spread of diseases like tuberculosis and also enhance the preparedness in terms of drug production.