Analysis and Prediction of Tuberculosis using Machine Learning Classifiers

Abstract

This paper deals with the deployment and evaluation of machine learning classifiers for prediction of tuberculosis. This research paper deploys five key machine learning classifiers Naive Bayes, Support Vector Machine, Decision Tree, K Nearest Neighbors and Random Forest. It is clearly understood that Support Vector Machine provides the best accuracy 99.3 % for the prediction of Pulmonary Tuberculosis (PTB) and Extrapulmonary Tuberculosis (EPTB) when compared with all other machine learning classifiers on Tuberculosis data set. An important challenge in machine learning is to build accurate and competent machine learning classifiers. Hence Support Vector Machine is a best suited Machine Learning Classifier for prediction of the PTB and EPTB.