We investigate whether it is possible to learn rule sets efficiently in a network structure with a single hidden layer using iterative refinements over mini-batches of examples. A first rudimentary version shows an acceptable performance on all but one dataset, even though it does not yet reach the performance levels of Ripper.
@article{arxiv.2106.10202,
title = {An Investigation into Mini-Batch Rule Learning},
author = {Florian Beck and Johannes Fürnkranz},
journal= {arXiv preprint arXiv:2106.10202},
year = {2021}
}