When More Data Hurts: Prioritizing Data Quality in ML
Collecting more data is often the instinct, but more data with poor labels can make models worse. This post explains how to audit data, detect label noise, and decide when to prioritize quality before scaling.