Towards understanding the importance of time-series features in automated algorithm performance prediction
收藏NIAID Data Ecosystem2026-03-13 收录
下载链接：
https://zenodo.org/record/6637636
下载链接
链接失效反馈官方服务：
资源简介：
merged_feature_importance.csv - CSV with feature importance values with different meta-models, forecasting algorithms, and feature importance methods computed on 30 different train/test splits.
Catch22.csv - Catch22 features (raw time-series)
Catch22Log.csv - Catch22 features (log time-series)
Catch22Diff.csv - Catch22 features (differenced time-series)
TSFresh.csv - TSFresh features (raw time-series)
TSFreshLog.csv - TSFresh features (log time-series)
TSFreshDiff.csv - TSFresh features (differenced time-series)
mape.csv - sMAPE performance for all forecasting algoirthms
创建时间：
2022-06-23
