Mathematical modeling and neural network based fitting of HIV/AIDS data in the workingclass population case study from Ethiopia.

HIV/AIDS deeply affects society, forcing people to leave their jobs and family responsibilities. This not only impacts their lives but also has a negative effect on the economy. HIV/AIDS affects development and growth by lowering productivity, income, and poverty levels. This study examines the impact of HIV/AIDS on Ethiopia's workingclass population using mathematical and neural network-based modeling, aiming to develop an accurate predictive framework. A supervised DNN with sigmoid activation is trained on WHO data (2000-2023) to predict disease incidence and transmission. Numerical simulations show that utilizing feedforward DNN methods gives precise solutions for advanced epidemiological models. The model attained a prediction accuracy above 99% and revealed a gradual rise in HIV incidence among nonproductive individuals over two decades. Finally, the trend of HIV/AIDS infection in Ethiopia from 2024 to 2050 is forecasted. The full-blown AIDS and infected class population approaches zero. Based on our estimated parameters and numerical simulations, it is evident that the application of neural network methods marks a significant advancement in the field of epidemiological modeling.