Forecast of Covid-19 Using Deep Learning

Outbreak of Covid-19 pandemic caused significant mortality and it became a threat to the human life. Since then every government is doing its best to curtail this health emergency. In India also the government has adapted a number of policies to contain this public health emergency. Despite of all these measures undertaken early prediction and forecast of the number of cases can greatly augment with the health policies implemented by the government bodies. In this direction various deep learning algorithms such as variants of Long Short Term Memory (LSTM) models are used in this study to forecast the number of cases. It is inferred from this study that the Bidirectional LSTM provided highest performance by providing a minimum mean absolute percentage error (MAPE) of 0.021% on the Indian Covid-19 data. The proposed methodology can be used in efficient planning and management of Covid-19.