Comparison of model performance based on a 10-fold cross validation (CV) between random forest (RFreg), gradient boost (GBreg), vector delay, flat delay, and basic linear regression (linReg) shown as box plot for the performance metrics (a) RMSE, (b) mean absolute error (MAE) and (c) mean error . Each box contains 10 folds from the cross-validation, its mean is shown in green, the median in orange, the box edge shows the 25/75 percentile and the whiskers indicate the full range.