Add SeFT as a classification model

[WenjieDu]

1. Model description

SeFT as a classification model for POTS should be included in PyPOTS.

@InProceedings{pmlr-v119-horn20a,
title = {Set Functions for Time Series},
author = {Horn, Max and Moor, Michael and Bock, Christian and Rieck, Bastian and Borgwardt, Karsten},
booktitle = {Proceedings of the 37th International Conference on Machine Learning},
pages = {4353--4363},
year = {2020},
editor = {Hal Daumé III and Aarti Singh},
volume = {119},
series = {Proceedings of Machine Learning Research},
month = {13--18 Jul},
publisher = {PMLR},
}

2. Check open-source status

• The model implementation is publicly available

3. Provide useful information for the implementation

https://github.com/BorgwardtLab/Set_Functions_for_Time_Series

[Justin0388]

I'm going to add this classification model to PyPOTS.

[WenjieDu]

I'm going to add this classification model to PyPOTS.

Many thanks, Jiaxin! I'll help review your PR code.