An efficient implementation of the Scalable Bayesian Rule Lists algorithm, a competitor to decision tree algorithms. The model builds from pre-mined association rules and has a logical structure identical to a decision list. The algorithm, developed by Hongyu Yang, Cynthia Rudin, and Margo Seltzer in 2017, is fully optimized over rule lists to balance accuracy, interpretability, and computational speed.
Use Cases
- Building interpretable binary classifiers based on pre-mined association rules.
- Comparing model performance against decision tree algorithms using the logical structure of a decision list.
- Training models where computational speed and accuracy must be balanced with interpretability.
Strengths
- Algorithm is fully optimized over rule lists, suggesting a deliberate design for efficiency.
- Model is described as striking a practical balance between accuracy, interpretability, and computational speed.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count and dataset scale are unknown, which may limit suitability assessment.
- Last update date is unknown; freshness unverified.
Provenance
- Source
- Implementation associated with the 2017 paper by Hongyu Yang, Cynthia Rudin, Margo Seltzer.
- Collection Method
- Algorithm implementation for building rule lists from pre-mined association rules.
- Time Range
- Algorithm published in 2017.
- Freshness
- Last updated date is unknown.
- Geography
- null