Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models
Abstract HIV (Human Immunodeficiency Virus) is a virus that causes the immune system to be damaged, thereby reducing the body’s ability to defend against infections and illnesses. In the absence of proper treatment, HIV can culminate into AIDS (Acquired Immunodeficiency Syndrome). The first-line app...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-10882-6 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849333418335141888 |
|---|---|
| author | Abdullah M. Almarashi |
| author_facet | Abdullah M. Almarashi |
| author_sort | Abdullah M. Almarashi |
| collection | DOAJ |
| description | Abstract HIV (Human Immunodeficiency Virus) is a virus that causes the immune system to be damaged, thereby reducing the body’s ability to defend against infections and illnesses. In the absence of proper treatment, HIV can culminate into AIDS (Acquired Immunodeficiency Syndrome). The first-line approach to HIV infection consists of antiretroviral therapy (ART), a combination of drugs that restrict virus replication. Effective prediction of infectious diseases is particularly vital for timely interventions and allocation of resources for disease management and prevention. This study focuses on identifying effective time series forecasting models for HIV and anti-retroviral therapy (ART) cases in Pakistan. The study utilized monthly reported HIV and ART cases data from the National AIDS Control Program, sourced from the Pakistan Bureau of Statistics, spanning the period from 2016 to 2021. Various time series models including ARIMA (Auto-regressive integrated moving average), exponential smoothing (Brown, Holt, Winter), neural network auto-regressive model (NNAR), and ETS (Exponential Smoothing State space) models were applied to analyze and forecast the monthly patterns of HIV and ART cases. Descriptive and time series analyses were conducted using the R programming language. The models were evaluated based on their ability to accurately capture and predict the fluctuations in HIV and ART cases over time. The average monthly cases for HIV and ART were found to be 36,405 ± 12,740 and 28,287 ± 12,485, respectively. Among the models evaluated, the NNAR (1,1,2) forecasting model emerged as the most accurate for both HIV and ART cases. It outperformed other competing models based on well-known accuracy measures such as RMSE, MAE, and MAPE. According to the selected NNAR(1,1,2) model, the study predicts a monthly increase of 4.98% in HIV cases and 16.32% in ART cases. The results proposed the non-linear approach of NNAR model to predict the AIDS and ART cases which help policymakers and healthcare professionals involved in disease management and prevention strategies in Pakistan to improve the policies and their implementation. |
| format | Article |
| id | doaj-art-b2c2b61d03f34ee5b729c331e8f1b0b1 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-b2c2b61d03f34ee5b729c331e8f1b0b12025-08-20T03:45:52ZengNature PortfolioScientific Reports2045-23222025-07-0115111710.1038/s41598-025-10882-6Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning modelsAbdullah M. Almarashi0Department of Statistics, Faculty of Science, King Abdulaziz UniversityAbstract HIV (Human Immunodeficiency Virus) is a virus that causes the immune system to be damaged, thereby reducing the body’s ability to defend against infections and illnesses. In the absence of proper treatment, HIV can culminate into AIDS (Acquired Immunodeficiency Syndrome). The first-line approach to HIV infection consists of antiretroviral therapy (ART), a combination of drugs that restrict virus replication. Effective prediction of infectious diseases is particularly vital for timely interventions and allocation of resources for disease management and prevention. This study focuses on identifying effective time series forecasting models for HIV and anti-retroviral therapy (ART) cases in Pakistan. The study utilized monthly reported HIV and ART cases data from the National AIDS Control Program, sourced from the Pakistan Bureau of Statistics, spanning the period from 2016 to 2021. Various time series models including ARIMA (Auto-regressive integrated moving average), exponential smoothing (Brown, Holt, Winter), neural network auto-regressive model (NNAR), and ETS (Exponential Smoothing State space) models were applied to analyze and forecast the monthly patterns of HIV and ART cases. Descriptive and time series analyses were conducted using the R programming language. The models were evaluated based on their ability to accurately capture and predict the fluctuations in HIV and ART cases over time. The average monthly cases for HIV and ART were found to be 36,405 ± 12,740 and 28,287 ± 12,485, respectively. Among the models evaluated, the NNAR (1,1,2) forecasting model emerged as the most accurate for both HIV and ART cases. It outperformed other competing models based on well-known accuracy measures such as RMSE, MAE, and MAPE. According to the selected NNAR(1,1,2) model, the study predicts a monthly increase of 4.98% in HIV cases and 16.32% in ART cases. The results proposed the non-linear approach of NNAR model to predict the AIDS and ART cases which help policymakers and healthcare professionals involved in disease management and prevention strategies in Pakistan to improve the policies and their implementation.https://doi.org/10.1038/s41598-025-10882-6Anti-retroviral therapyForecastingHIVPakistanTime series models |
| spellingShingle | Abdullah M. Almarashi Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models Scientific Reports Anti-retroviral therapy Forecasting HIV Pakistan Time series models |
| title | Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models |
| title_full | Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models |
| title_fullStr | Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models |
| title_full_unstemmed | Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models |
| title_short | Statistical modelling and forecasting of HIV and anti-retroviral therapy cases by time-series and machine learning models |
| title_sort | statistical modelling and forecasting of hiv and anti retroviral therapy cases by time series and machine learning models |
| topic | Anti-retroviral therapy Forecasting HIV Pakistan Time series models |
| url | https://doi.org/10.1038/s41598-025-10882-6 |
| work_keys_str_mv | AT abdullahmalmarashi statisticalmodellingandforecastingofhivandantiretroviraltherapycasesbytimeseriesandmachinelearningmodels |