<![CDATA[The data shows that HIV (Human Immunodeficiency Virus) has caused tens of millions of global deaths, with 630,000 people dying from HIV-related illnesses in 2022 and 1.3 million people newly infected with HIV. Without treatment, HIV can progress to AIDS (Acquired Immune Deficiency Syndrome), weakening the immune system and increasing the risk of infections and other diseases. Despite advancements in treatment, early detection of AIDS remains a priority. This research develops an AIDS prediction model using machine learning, which proves to be an effective solution in providing future health predictions. However, data imbalance issues challenge the model in predicting rare AIDS cases. To solve this problem, oversampling techniques are employed to balance the distribution of minority classes. This study explores oversampling techniques such as SMOTE, ADASYN, and Random Oversampling, combined with the XGBoost algorithm. The results show that the combination of Random Oversampling technique with the XGBoost Classifier yields the best performance with an accuracy of 94.44%, precision of 90.72%, recall of 98.74%, and an f1_score of 94.65%. This research is expected to provide valuable insights for healthcare practitioners and the public in efforts to control the spread of AIDS globally.]]>
Copyrights © 2024
