|
| Ref. | Method | Application | Results obtained | Metrics used | Data unit |
|
| [40] | MTS-LSTM | Healthcare COVID-19 prediction | Excellent spatial granularity superior predicted performance many weeks ahead. | RMSE | Daily |
| [41] | TEG-net | Healthcare (heart rate) | TEG-net exceeds the second-best baseline in terms of area under the receiver operating characteristic curve and area under the precision-recall curve. | ROC-AUC, PR-AUC, F1-score, precision, recall, accuracy | Every minute |
| [42] | LSTM | Healthcare | It is possible to estimate vital signs with a high degree of accuracy. | Accuracy MAE >80% | Hourly |
| [43] | Stacked LSTM, LSTM | Healthcare (forecasting pandemic) | The stacked LSTM models predicted more reliably and generated good results. | MAPE | Daily |
| [44] | M-LSTM | Healthcare (forecasting pandemic) | The greatest result was discovered for predicting the pandemic’s future trajectory | MAPE, RMSE, NRMSE, and R2 are 0.51, 458.12, 0.0016, and 0.9 | Daily |
| [45] | LSTM | Healthcare (Parkinson’s disease diagnosis) | Adam-optimized LSTM networks are capable of successfully learning gait kinematic data and provide an average accuracy | Accuracy improvement of 3.4% | Every 2 minutes |
| [46] | Stacked, bidirectional, and convolutional LSTM | Healthcare COVID-19 prediction | Convolutional LSTM outperformed the other two models | MAPE, accuracy, precision, recall, F-measure | Daily |
| [47] | LSTM-GA | Influenza outbreak forecasting | The experimental results show that the hybrid model presented here outperforms other well-developed machine learning approaches | RMSE | Weekly |
| [48] | ANN | Healthcare COVID-19 prediction | The developed ANN model for predicting COVID-19 cases was found to be stable | MAPE and RMSE | Daily |
| [49] | GRU | COVID-19 prediction | GRU can achieve better convergence and high performance with the simplest structure | R2 | Daily |
|
