HIV Status Predictive Modeling Using Data Mining Technology

This research work has attempted to investigate the underlying determinant factors of being HIV positive or HIV negative from the available HCT data using data mining techniques. It consists of experiments on searching classification model that predicts HIV status and association rule mining to discover the relationship of HIV status with the selected attributes. The classification experiments are carried out using J48 and ID3 algorithms. The association rule mining is using Apriori algorithm. One of the surprising results obtained from the experiment was that age group 50 and above are becoming also vulnerable to HIV/AIDS as the patterns have indicated. Medical experts also have stated that a growing number of older people are being infected with HIV/AIDS. One of the reasons they raised is that, they are finding HIV more often than ever before in older age since improved treatments are helping people with the disease live longer. The second reason is doctors do not always test older people for HIV/AIDS and so may miss some cases during routine check-ups and others.

Autorentext

I have M.Sc in Health Informatics from Addis Ababa university. Now I am working at Adama Science and Technology University as a Lecturer.