Customer age and purchase amount contain missing values.
Do not silently treat missingness as a meaningful value.
Choose two preprocessing practices.
Investigate why, where, and when values are missing, fit imputation on training data, and evaluate missing indicators and distribution effects.
Detailed explanation
Leakage and missing patterns are controlled.
Missingness may be random or operationally meaningful.
Zero can be misinterpreted and create bias.
Meaning and inference conditions diverge.
Check official documentation
Compare the following points with the official documentation.
Review Task 1.3 and the preprocessing and leakage guidance. For missing age or purchase amounts, describe reasons, rates, and group differences, then outline applying training-fitted imputation and missingness indicators to validation data.
Official references
What to confirm
Explain why imputation statistics are fitted on training data only and why missingness patterns should also be investigated.
Key points
- Missingness
- Imputation
- Leakage
Notes
- Environment: Reference review of official documentation and hypothetical examples; no AWS resources or model training are required.
- Distinguish a genuine business value of zero from a failed or unavailable measurement. Do not assume an imputed value means the same thing as an observed zero.
- Evaluate whether a missingness indicator helps. Logging or workflow changes can alter missingness patterns, so continue checking them in production.
Foundation review
Missing value
A required or expected attribute that has no recorded value for a data item.
Imputation
Replacing a missing value with a rule- or model-derived estimate.