Analysis of Tuberculosis Disease Using Association Rule Mining

Abstract

Tuberculosis (TB) is a chronic infectious disease and remains a serious explanation for death globally. Most of the infected people are from poverty-stricken communities with low healthcare infrastructure. Machine learning (ML) shows a unique way to facilitate the diagnosis of TB. ML can provide deep insights into large biomedical datasets, as well as it can uncover new biomedical and healthcare knowledge. ML techniques have to be utilized to successfully recognize TB occurrence at an early stage or re-occurrence. Association rule mining and decision tree can be used for prediction of the occurrence and re-occurrence of TB. Association rules and decision tree give a meaningful and efficient way to define and present certain dependencies among the attributes in a dataset ML algorithms help to identify, discover hidden and valuable information, and have been massively used and explored in the medical domain. The main purpose of this study is to apply ML techniques for extracting hidden patterns, which are significant to predict TB at an early stage.