Tuberculosis and Lung Cancer Prediction using Machine Learning Methods and Over-Sampling Technique

Abstract

With the continuous advancement of technology, people and machines can complement their specific skills to achieve effective results. In this sense, with the inevitable increase in the number of diseases that threaten human health, Decision Support Systems (DSS) are widely used in the medical field to help doctors making better clinical decisions. Among these diseases, such as tuberculosis and lung cancer are considered potentially serious infectious and are among the top 10 causes of death in the world. This paper presents a medical DSS for tuberculosis and lung cancer diagnosis by using machine learning algorithms, such as the Support Vector Machines (SVM) and Artificial Neural Network (ANN). Moreover, Borderline Synthetic Minority Over-Sampling Technique (Borderline - SMOTE) was also employed to increase the number of minor sample size. The experimental dataset used is taken from Diyarbakir chest diseases hospital. The obtained results proved the efficiency of the proposed system in helping doctors making the right decision and improving the quality of health care.