%0 Generic %1 loning2020forecasting %A Löning, Markus %A Király, Franz %D 2020 %K py:lib sktime timeseriesforecasting todo:read %T Forecasting with sktime: Designing sktime's New Forecasting API and Applying It to Replicate and Extend the M4 Study %U http://arxiv.org/abs/2005.08067 %X We present a new open-source framework for forecasting in Python. Our framework forms part of sktime, a more general machine learning toolbox for time series with scikit-learn compatible interfaces for different learning tasks. Our new framework provides dedicated forecasting algorithms and tools to build, tune and evaluate composite models. We use sktime to both replicate and extend key results from the M4 forecasting study. In particular, we further investigate the potential of simple off-the-shelf machine learning approaches for univariate forecasting. Our main results are that simple hybrid approaches can boost the performance of statistical models, and that simple pure approaches can achieve competitive performance on the hourly data set, outperforming the statistical algorithms and coming close to the M4 winner.

@misc{loning2020forecasting, abstract = {We present a new open-source framework for forecasting in Python. Our framework forms part of sktime, a more general machine learning toolbox for time series with scikit-learn compatible interfaces for different learning tasks. Our new framework provides dedicated forecasting algorithms and tools to build, tune and evaluate composite models. We use sktime to both replicate and extend key results from the M4 forecasting study. In particular, we further investigate the potential of simple off-the-shelf machine learning approaches for univariate forecasting. Our main results are that simple hybrid approaches can boost the performance of statistical models, and that simple pure approaches can achieve competitive performance on the hourly data set, outperforming the statistical algorithms and coming close to the M4 winner.}, added-at = {2021-03-25T15:38:15.000+0100}, author = {Löning, Markus and Király, Franz}, biburl = {https://www.bibsonomy.org/bibtex/2771a1ce3207fab5dc80e60f8d8c1e9b6/annakrause}, description = {[2005.08067] Forecasting with sktime: Designing sktime's New Forecasting API and Applying It to Replicate and Extend the M4 Study}, interhash = {b2ad9b8eb534f3bd6328ad3d3f96034a}, intrahash = {771a1ce3207fab5dc80e60f8d8c1e9b6}, keywords = {py:lib sktime timeseriesforecasting todo:read}, note = {cite arxiv:2005.08067}, timestamp = {2021-12-16T17:39:13.000+0100}, title = {Forecasting with sktime: Designing sktime's New Forecasting API and Applying It to Replicate and Extend the M4 Study}, url = {http://arxiv.org/abs/2005.08067}, year = 2020 }