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

Abstract

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 Bangladesh, Brazil, India, 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. Importantly, CNN model showed clear indications of imminent second wave of COVID-19 in the above-mentioned countries. 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 results enlighten some surprising results that the studied countries are going to witness dreadful consequences of second wave of COVID-19 in near future. Quick responses from government officials and public health experts are required in order to mitigate the future burden of the pandemic in the above-mentioned countries.

Feature extraction with capsule network for the COVID-19 disease prediction though X-ray images

“…Finally the last capsule layer with size 1x2 is applied for the prediction of the disease. Batch Normalization – Batch normalization is applied to form the mini batch of the images [15] . Max pooling – This technique is used for the discretization purpose which helps in dimensionality reduction and more informative features are generated [16] .…”

A predictive analytics model for COVID-19 pandemic using artificial neural networks

“…Kolozsvári et al [31] explored the use of artificial intelligence in modeling COVID-19. Nabi et al [32] studied four deep learning models including LSTM, GRU, CNN and MCNN to predict COVID-19 cases. Namasudra et al [33] proposed a new Nonlinear Autoregressive (NAR) Neural Network Time Series (NAR-NNTS) model to predict COVID-19 cases.…”

Outbreak prediction of COVID-19 using Recurrent neural network with Gated Recurrent Units

“…Researchers carried out a rigorous comparison of the long-term memory, Convolutional Neural Networks (CNN) [7] , Gated Recurring Unit (GRU) [8] , and Multivariate Convolutional Neural Network (MCNN) , alongside an analysis of the projections on the number of new COVID-19 cases daily for Brazil, Russia, and the United Kingdom for the four deep learning methods: short-term memory, CNN, Gated Recurring Unit (GRU), and MCNN. The robustness of the model was tested using mean absolute per cent error (MAPE) and standard root mean square error (nRMSE) [9] . This shows that CNN is a superior deep learning model than the competition.…”