A complimentary set of tasks to the AutoML benchmark that can be used as a training set for meta-learning as suggested by Feurer et al. in the paper "Auto-Sklearn 2.0: Hands-free AutoML via Meta-Learning".
A full description of the inclusion criteria can be found at in the paper at https://arxiv.org/abs/2007.04074
If you use this work in a publication please cite:
@article{feurer-arxiv21a,
title = {Auto-Sklearn 2.0: Hands-free AutoML via Meta-Learning},
author = {Matthias Feurer and Katharina Eggensperger and Stefan Falkner and Marius Lindauer and Frank Hutter},
journal = {arXiv:2007.04074 [cs.LG]},
year = {2021}
}