Enhancing COVID-19 forecasting through deep learning techniques and fine-tuning

Alba Puelles López, Rodrigo Martínez-Béjar, Kusrini Kusrini, Arief Setyanto, I Made Artha Agastya


In this study, a comprehensive analysis of classical linear regression forecasting models and deep learning techniques for predicting coronavirus disease of 2019 (COVID-19) pandemic data was presented. Among the deep learning models, the long short-term memory (LSTM) neural network demonstrated superior performance, delivering accurate predictions with minimal errors. The neural network effectively addressed overfitting and underfitting issues through rigorous tuning. However, the diversity of countries and dataset attributes posed challenges in achieving universally optimal predictions. The current study explored the application of the LSTM in predicting healthcare resource demand and optimizing hospital management to provide potential solutions for overcrowding and cost reduction. The results showed the importance of leveraging advanced deep learning techniques for improved COVID-19 forecasting and extending the application of the models to address broader healthcare challenges beyond the pandemic. To further enhance the model performance, future work needed to incorporate additional attributes, such as vaccination rates and immune percentages.


Coronavirus; Deep learning; Enhanced prediction; Fine-tuning; Forecasting;

Full Text:


DOI: http://doi.org/10.11591/ijece.v14i1.pp934-943

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

International Journal of Electrical and Computer Engineering (IJECE)
p-ISSN 2088-8708, e-ISSN 2722-2578

This journal is published by the Institute of Advanced Engineering and Science (IAES) in collaboration with Intelektual Pustaka Media Utama (IPMU).