ML 8.5Lesson outlineLesson 5 of 7 in Module 8
Data Leakage and Pipelines
Find the ways test or future information gets into training, and prevent them with a pipeline.
Video in production.
Lesson outline: Data Leakage and Pipelines
In this lesson10 parts
- 01Scores that do not survive
- 02The calls, again
- 03A column that knows the answer
- 04A step fitted on every row
- 05A shuffle of time
- 06Time-ordered folds
- 07The pipeline
- 08Searching the pipeline
- 09Where it misleads
- 10Code, recap, and what comes next
Builds on
This lesson stands on its own. These go deeper into what it uses.
- ML 2.2: The Modelling Checks: Target, Leakage and Feature Decisions
- 4.4
- 4.5
- 5.3
- 5.4
- 6.1
- 8.1–8.4