An Investigation into Mini-Batch Rule Learning
arXiv:2106.10202v13 citations
Originality Synthesis-oriented
AI Analysis
This is an incremental improvement for rule learning in machine learning, focusing on efficiency in network structures.
The paper tackled the problem of learning rule sets efficiently in a neural network with a single hidden layer using mini-batches, showing acceptable performance on most datasets but not matching Ripper's levels.
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.