Predicting Mortality in COVID-19 Patients Based on Symptom Data Using Hybrid Neural Networks

Accurate prediction of COVID-19 cases is crucial for effective public health planning and resource allocation. In this study, we evaluated the performance of different neural network architectures, including LSTM, RNN, ANN, and hybrid models for COVID-19 case prediction. Our results demonstrate that the hybrid models, combining the strengths of LSTM, RNN, and ANN architectures, outperform the individual models in terms of accuracy, precision, recall, and F1-score. The best-performing model, LAR, achieves an accuracy of 95.25%, precision of 97%, recall of 96.68%, and F1-score of 96.84%. Our findings suggest that the hybrid approach can significantly improve the accuracy of COVID-19 case prediction, which could have important implications for public health policy and decision-making. Future work could focus on developing more sophisticated hybrid models to further improve prediction accuracy.