Data Analysis for ML/AI
Labels and data quality
Label noise, class imbalance, dataset versioning, and audit-grade quality.
5 lessons
Follow in order
Use the button below to sign in and unlock lessons.
Your path
Lessons in sequence
Work through these in order—each lesson builds on the previous one.
-
Lesson 1 of 5
Label noise: where dirty labels come from
-
Lesson 2 of 5
Gold sets and inter-annotator agreement
-
Lesson 3 of 5
Weak supervision and programmatic labels
-
Lesson 4 of 5
Label drift and concept change
-
Lesson 5 of 5
Label quality audits before training