• Login
    View Item 
    •   NM-AIST Home
    • Computational and Communication Science Engineering
    • Masters Theses and Dissertations [CoCSE]
    • View Item
    •   NM-AIST Home
    • Computational and Communication Science Engineering
    • Masters Theses and Dissertations [CoCSE]
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Machine learning model for prediction and visualization of HIV index testing in northern Tanzania

    Thumbnail
    View/Open
    Full text (1.518Mb)
    Date
    2022-07
    Author
    Chikusi, Happyness
    Metadata
    Show full item record
    Abstract
    Infection with the human immunodeficiency virus and acquired immunodeficiency syndrome (HIV/AIDS) continue to pose a threat to Tanzanian society. Various tactics have been used to improve the number of persons who are aware of their HIV status. Index testing stands out among these methods as the most effective way to count the number of HIV contacts who may be at risk of catching HIV from HIV-positive individuals. The current HIV index testing, however, is manual, which presents a number of difficulties, including inaccuracies, is time consuming, and is expensive to operate. In order to forecast and depict HIV index testing, this study presents the findings of the machine-learning model. The software development procedure was in accordance with agile software development principles. The regions of Kilimanjaro, Arusha, and Manyara in Tanzania are where the data was gathered which consisted of 11 features and 6346 samples. The dataset was then separated into training sets with 5075 samples each and testing sets with 1270 samples (80/20). The datasets were subjected to the methods Random Forest (RF), XGBoost, and Artificial Neural Networks (ANN). Random forest MAE (1.1261), XGBoost MAE (1.2340), and ANN MAE (1.1268) were the three results obtained. Random forest algorithms had the lowest mean absolute errors (MAE). Therefore, RF appearing to have the highest performance when compared to the other two algorithms. In comparison to men (17.4%), data visualization reveals that females are more likely to test for HIV and to name their partners (82.6%). Additionally, there were higher instances of persons listing and mentioning their partners in the Kilimanjaro region. This work helped us realize the importance of machine learning in predicting and visualizing HIV index tests in general. The created model can help decision-makers build a viable intervention to stop the spread of HIV and AIDS in our communities. The report suggests that health centers in other areas employ this concept to make their work more straightforward.
    URI
    https://doi.org/10.58694/20.500.12479/1591
    Collections
    • Masters Theses and Dissertations [CoCSE]

    Nelson Mandela-AIST copyright © 2021  DuraSpace
    Theme by 
    Atmire NV
     

     

    Browse

    All PublicationsCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

    My Account

    LoginRegister

    Nelson Mandela-AIST copyright © 2021  DuraSpace
    Theme by 
    Atmire NV