Deep learning time series to forecast COVID-19 active cases in INDIA: A comparative study

Abstract

In the present situation of "COVID-19 pandemic"which devastated worldwide socioeconomic implications that led by Indian government to initiate and to perform intense procedures to control the spread and impact by dispensing the capability to predict the outbreak, which hits the crest to decrease the impact of disease and guiding the government to update its policies as required for implementing protective steps needed for "Public Health System (PHS)". These methodologies tend to be transforming among people for improvisation of vigor system. In this paper, we investigate thoroughly the precision of various "Time series"modeling techniques for detecting "corona virus outbreak"in topmost ten different states with the maximum number of "confirmed cases"as on 31 August 2020. We implemented six different deep learning approaches on time series for comparing the values associated in datasets that relates to the progression of the virus in each state based on the population as the attained results represent the scaling of time series methods accurately and predict various affected aspects in near future.