history_size/window_sampling_limit is shorter than input size

[gorold]

For certain datasets, e.g. Yearly/Quarterly/Monthly M4 datasets, the quantity history_size is set to 1.5, leading window_sampling_limit to be 1.5 times of the horizon length. Yet, the input size could be up to 7 times the horizon length, meaning that during training phase the model mostly observes padding. Is this an issue, and possibly leading to a degradation in performance in these dataset?

[acifonelli]

Leaving the comment here for the future. As stated by the authors in the paper - Appendix D.1, page 26 on arxiv -

Target and history points that fall outside of the time series limits given the anchor position are filled with zeros and masked during the training

Hence there should be no performance degradation.