Do Rule-Derived Labels Improve Prediction When Reported Labels Are Limited? A Two-Site Machine Learning Study
**Background.** Rule derived (weak) labels can enlarge a training set when reported labels are scarce. Their value depends on label quality, on how the ext