Forecasting COVID-19 cases: A comparative analysis between Recurrent and Convolutional Neural Networks

When the entire world is waiting restlessly for a safe and effective COVID-19 vaccine that could soon become a reality, numerous countries around the globe are grappling with unprecedented surges of new COVID-19 cases. As the number of new cases is skyrocketing, pandemic fatigue and public apathy towards different intervention strategies are posing new challenges to the government officials to combat the pandemic. Henceforth, it is indispensable for the government officials to understand the future dynamics of COVID-19 flawlessly in order to develop strategic preparedness and resilient response planning. In light of the above circumstances, probable future outbreak scenarios in Brazil, Russia and the United kingdom have been sketched in this study with the help of four deep learning models: long short term memory (LSTM), gated recurrent unit (GRU), convolutional neural network (CNN) and multivariate convolutional neural network (MCNN). In our analysis, CNN algorithm has outperformed other deep learning models in terms of validation accuracy and forecasting consistency. It has been unearthed in our study that CNN can provide robust long term forecasting results in time series analysis due to its capability of essential features learning, distortion invariance and temporal dependence learning. However, the prediction accuracy of LSTM algorithm has been found to be poor as it tries to discover seasonality and periodic intervals from any time series dataset, which were absent in our studied countries. Our study has highlighted the promising validation of using convolutional neural networks instead of recurrent neural networks when it comes to forecasting with very few features and less amount of historical data.