Introduction:
At present, especially in developing countries, human immunodeficiency virus/acquired immune deficiency syndrome become very extreme. As we know, the virus break up and damages the function of immune system, and infected people gradually become immunodeficient. Artificial intelligence play a great role in the prediction of diseases. The function of immune system is the level of CD4+ cell count. In this study, ensemble method was applied.

Material and methods:
Records used in the study were collected from world data available online. Random forest, naïve Bayes, J48, and KNN algorithms as well as the proposed ensemble method using bagging technique were employed to fit the rate data, with 10k cross-validation. The performance of the model was evaluated using accuracy, precision, sensitivity, and F-score.

Results:
Based on the experiment, the proposed ensemble method achieved the best results in metrics of accuracy, precision, and F-score compared with other models by providing accuracy of 95.36%, precision of 90.23%, recall of 91.5%, and F-score of 92.35%.

Conclusions:
Currently, machine learning plays a great role in the prediction of future, especially in health industry. The power of models is depending on the data. Based on the results of the study, the proposed ensemble method provided the best accuracy compared with other machine learning methods, with accuracy of 95.36% shown as very effective result.