Changed the dataset so that faulty values of new_cases_(smoothed) are not included. Changed Hyperparameters accordingly

Changed Visualizations of Regression to ribbon, changed the visualization of the Feature Importance so that it only shows relative values

changed the data so that no inddependent death variable is used now. Retrained models accordingly. Added a graph visualizing feature importance

Added Visualization of the Predicted and Measured Values of the target variable in a Scatter Plot for each Model

Added Slovakia to data, retrained every model with the new dataset added with changed optimal hyperparameters

Changed calculation of Time so it only measures fitting and prediction now takes the average for the folds for each run, use CPU instead of GPU for LGBM,  use sample size 1200

Replaced normal Cross Validation with ShuffleSplit CV in order to obtain different Splits for every call