What do you monitor about features (not just models)?

medium

Answer

$(1) Freshness: is the feature updated on schedule?$
$(2) Completeness: missing rate per feature.$
$(3) Distribution drift: PSI / MMD vs baseline.$
$(4) Cardinality: for categorical, new unseen values.$
$(5) Correlation stability: pairwise correlations vs baseline.$
$(6) Feature importance stability: SHAP over time.$
$(7) Range violations: values outside expected.$
$Store per-feature health score \to dashboard.$
$Predictive: feature health degrades before model does.$

Check yourself — multiple choice

Random
Freshness + completeness + drift (PSI/MMD) + cardinality (new values) + correlation stability + SHAP importance stability + range violations; feature health degrades before model
Only models
Not real

Feature monitoring: freshness + completeness + drift + cardinality + correlation + importance + range.

#monitoring#features

Practise MLOps & Data Quality

215 interview questions in this topic.

Read the lesson Flashcards Take the quiz

Related questions