Forecasting the Spread and Mortality Trends of COVID-19 in Afghanistan and Iran Using Long Short-Term Memory Neural Networks
Salamat Academic & Research Journal · 2025
DOI: 10.61438/sarj.v1i2.145 ↗Kurz gefasst
Fachartikel in der Zeitschrift Salamat Academic & Research Journal, 2025. Verfasst von Mohammad Jafari. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Objectives
This study aims to investigate and compare the prediction of COVID-19 prevalence and mortality trends in Iran and Afghanistan using the Long Short-Term Memory (LSTM) neural network model.
Methods
Daily epidemiological data from Iran and Afghanistan, spanning 2021 to 2023, were sourced from the Daily World database. The dataset was split into 80% for training and 20% for evaluation. The LSTM model was utilized to predict trends in COVID-19 cases and deaths for both countries.
Results
The LSTM model demonstrated high accuracy in forecasting COVID-19 trends in Afghanistan, achieving R² = 0.98, RMSE = 1.09, and MAE = 0.42. In contrast, the model's performance in Iran was markedly lower, with R² = -0.15, RMSE = 107.64, and MAE = 58.62. The discrepancy may stem from differences in data quality, greater volatility in the Iranian data, and variations in geographical and surveillance systems.
Conclusion
This research highlights the challenges and potential of LSTM models in forecasting COVID-19 trends in developing countries. The results underscore the need for refined approaches that address regional data complexities to enhance the accuracy and utility of predictive tools in health crisis management.
Zitieren
APA
Jafari, M. (2025). Forecasting the Spread and Mortality Trends of COVID-19 in Afghanistan and Iran Using Long Short-Term Memory Neural Networks. Salamat Academic & Research Journal. https://doi.org/10.61438/sarj.v1i2.145
Vancouver
Jafari M. Forecasting the Spread and Mortality Trends of COVID-19 in Afghanistan and Iran Using Long Short-Term Memory Neural Networks. Salamat Academic & Research Journal. 2025. doi:10.61438/sarj.v1i2.145
BibTeX
@article{Jafari2025,
author = {Jafari, Mohammad},
title = {Forecasting the Spread and Mortality Trends of COVID-19 in Afghanistan and Iran Using Long Short-Term Memory Neural Networks},
journal = {Salamat Academic & Research Journal},
year = {2025},
doi = {10.61438/sarj.v1i2.145}
}
EUVIMED