Predicting COVID-19 using lioness optimization algorithm and graph convolution network.

Area of Science:

• Epidemiology
• Data Science
• Computational Biology

Background:

• COVID-19 poses a significant public health challenge requiring accurate predictive models.
• Understanding disease transmission dynamics is crucial for effective public health interventions.
• Existing models may not fully capture the complex spatio-temporal patterns of infectious disease spread.

Purpose of the Study:

• To propose a novel prediction model, the Lioness Optimization Algorithm-Graph Convolutional Network (LsOA-GCN), for forecasting COVID-19 cumulative cases.
• To effectively integrate temporal and spatial features for enhanced epidemic prediction accuracy.
• To evaluate the performance of the LsOA-GCN model against established prediction methods.

Main Methods:

• Development of the LsOA-GCN model combining graph convolutional networks for spatial information and the Lioness Optimization Algorithm (LsOA) with Spearman correlation for temporal feature extraction.
• Application of the model to predict cumulative COVID-19 cases in 17 regions of Hubei Province.
• Rigorous evaluation using performance metrics and statistical tests, with comparisons to 10 other prediction methods.

Main Results:

• The LsOA-GCN model demonstrated superior performance across all evaluation indicators compared to 10 benchmark prediction methods.
• The model successfully captured essential spatio-temporal information from the feature data.
• Accurate prediction of epidemic trends in various regions of Hubei Province was achieved.

Conclusions:

• The LsOA-GCN model offers a significant advancement in COVID-19 cumulative case prediction.
• The integration of LsOA and GCN effectively models the complex spatio-temporal dynamics of infectious diseases.
• This approach provides a reliable tool for public health authorities to forecast and manage epidemic outbreaks.