THE IMPLEMENTATION OF PREDICTIVE MODEL TO DETECT FIRST-LINE ANTIRETROVIRAL THERAPY FAILURE AMONG HIV/AIDS PATIENTS IN NIGERIA
This study utilizes expert consultation to develop a machine learning-based predictive model that detects clients who are at high risk of treatment failure among those who are receiving first-line ARV therapy. The study uses retrospective cross-sectional data of clients who are at least 6 months on ART. The study has followed the Cio data mining model. The study has conducted two main procedures for model development; cluster modeling and classification modeling. The cluster modeling was conducted by using the K-mean algorithm and classification modeling was conducted by implementing decision tree (J48), NaiveBayes, SVM, and random forest algorithms The experimentation results show that all the algorithms were the same in terms of accuracy (98.998%), precision (0.990), recall (1.00), and F1-score (0.995). They differ in the time taken to build the classification model. J48 and Naïve Bayes algorithms have better time efficiency. Accordingly, the J48 and Naïve Bayes algorithms were found the best algorithms to develop an ART treatment detection model for the data considered in this study.
THE IMPLEMENTATION OF PREDICTIVE MODEL TO DETECT FIRST-LINE ANTIRETROVIRAL THERAPY FAILURE AMONG HIV/AIDS PATIENTS IN NIGERIA GET MORE COMPUTER SCIENCE PROJECT TOPICS AND MATERIALS