Towards understanding the importance of time-series features in automated algorithm performance prediction

<strong>merged_feature_importance.csv</strong> - CSV with feature importance values with different meta-models, forecasting algorithms, and feature importance methods computed on 30 different train/test splits. <strong>Catch22.csv</strong> - Catch22 features (raw time-series) <strong>Catch22Log.csv</strong> - Catch22 features (log time-series) <strong>Catch22Diff.csv</strong> - Catch22 features (differenced time-series) <strong>TSFresh.csv</strong> - TSFresh features (raw time-series) <strong>TSFreshLog.csv</strong> - TSFresh features (log time-series) <strong>TSFreshDiff.csv</strong> - TSFresh features (differenced time-series) <strong>mape.csv</strong> - sMAPE performance for all forecasting algoirthms