LSTM-based Sequential Time-Series Prediction of Covid-19 Cases

Project supervisor: Professor Adway Mitra, Center of Excellence in Artificial Intelligence, IIT Kharagpur (December 2020 – February 2021).

The project involves the analysis of the daily Covid-19 cases across 180 countries over one year to forecast rate of spread. The steps in our analysis involved:

• Categorising the phases of Covid-19 increase as Fast, Moderate and Slow using the Knee Detection Algorithm.
• Analysing the relationship between Phase change and NPIs (Non-pharmaceutical interventions) implemented by the government. Please find detailed analysis in the Phase-NPI Analysis Report.
• Implementing Phase-Specific LSTM Models to predict the rate of spread of Covid-19 in the fast, moderate and slow phases respectively, and compared the accuracy metrics of each model.