Abstract
The power of antiretroviral (ART) drugs has led to significant development in lowering and maintaining viral load, especially when early adherence to ART drugs is duly and successfully followed. Failure to adhere to treatment and guideline compliance leads to cell infection advancing to Acquired Immune Deficiency Syndrome (AIDS) and death. In this study, Supervised machine learning (ML) techniques and Exploratory data analysis were employed for data modelling. The Multilayer Perceptron (MLP), Support Vector Machine (SVM), Adaptive Neuro-Fuzzy Inference System (ANFIS), Multiple Linear Regression (MLR), and Binary Logistic Regression (BLR) were employed for ML prediction and class membership prediction. Multiple Correspondence Analysis (MCA) was used for exploratory data analysis to predict distance and assess model accuracy. The study utilized random sampling to select 1500 HIV patients on ART. The MLP machine learning model showed higher variability and lower error in training (R2 = 0.920209; MSE = 0.002150) and testing (R2 = 0.912092; MSE = 0.002283) than the other models.
Keywords
Antiretroviral therapy
AI models
Discrimination measures
Sensitivity
Viral load