Weather and population based forecasting of novel COVID-19 using deep learning approaches

Abstract

The spread of novel corona virus across the globe has a significant impact on various stake holders and posting a major challenge to the research community. Government has taken several measures for maintaining social distance and containment of disease, but still it is not a sufficient for the developing countries like India where the level of understanding the issue is deprived and hence it is a major challenge to the Health Care professionals. Therefore, it is mandatory that a prediction of the number of possible cases enables the preparedness of the Government and the Hospitals in resolving the issues and to take measures in controlling the spread of the disease Series. Deep learning model has been built by considering the features of weather and COVID-19 data (recovered, infected and deceased) for predicting the number of cases expected in India. The model is built on Concurrent Neural Network (CNN), Recurrent Neural Network (RNN), Bidirectional RNN (BRNN), Long Short-Term Memory (LSTM) and Bidirectional LSTM (BLSTM) based on the daily weather and COVID-19 data collected from Indian subcontinent. The results revealed that the algorithm BRNN yields a better prediction model when compared with the other models.