Using Machine Learning to Analyse the Effect of Antiretroviral therapy (ART) on People with HIV

Human immunodeficiency virus or HIV is a virus that attacks the body's immune system. If HIV is not treated it can develop into AIDS or acquired immunodeficiency syndrome. Even though HIV can not be cured permanently, Antiretroviral therapy or ART is believed to be able to reduce the spread of HIV and increase the life expectancy of the person who is infected by HIV. Machine Learning (ML) has proven to be useful for helping researchers in the medical field through data analysis. In our experiment, we collected data in 11 year period that consist of 75 countries with ART coverage, HIV case, and their total population. The Machine-Learning algorithms that we are going to use (1) K-Nearest Neighbor (KNN) algorithm, (2) Logistic Regression, (3) Support Vector Machine (SVM), (4) XGBoost, (5) Decision Tree, (6) Random Forest, and (7) K-Means clustering. The contribution of this paper is to explain our findings on predicting the Antiretroviral therapy effect on people with HIV. The optimum result of the supervised Machine Learning result in Random Forest with an accuracy of 90%. Meanwhile, the unsupervised Machine learning resulted in 2 clusters, the first cluster consists of 15 countries while the second cluster consists of 60 countries.