Tag
This paper evaluates traditional machine learning models versus LLM-generated rule-based systems for heart disease prediction, showing that traditional models have higher accuracy but LLM rules provide better interpretability for clinical use.