Learning Interpretable Rules for Scalable Data Representation and Classification. (arXiv:2310.14336v3 [cs.LG] UPDATED)

Rule-based models, e.g., decision trees, are widely used in scenarios
demanding high model interpretability for their transparent inner structures
and good model expressivity. However, rule-based models are hard to optimize,
especially on large data sets, due to their discrete parameters and structures.
Ensemble methods and fuzzy/soft rules are commonly used to improve performance,
but they sacrifice the model interpretability. To obtain both good scalability
and interpretability, we propose a new classifier, named Rule-based
Representation Learner (RRL), that automatically learns interpretable …

arxiv classification cs.lg data data sets decision decision trees ensemble good interpretability model interpretability parameters representation rules scalable trees