In this project, the time-series data of COVID-19 cases in Germany is analysed in 2 different scenarios (cumulative cases and daily new cases) with 2 different models to gain insights into future case numbers: (1) the decomposition model (Prophet), (2) the auto-regressive model (AR). Lastly, a LSTM (Long Short-Term Memory) recurent neural network (RNN) was used. Performances of these models can be seen in the plots.
Case numbers were taken from the COVID-19 case numbers for Germany repository.