Preskoči na glavni sadržaj
🔒 Način pregleda. Prvih petnaest Foundations lekcija je besplatno; ova je Pro. Pokrenite 7-dnevni trial da otključate editor, AI savjete i ostatak programa. Kartica obavezna, otkažite bilo kada u Dashboard.
Pokreni 7-dnevni trial →
⚡
← Kolegiji
›
Data Science Applied
›
Module 2 · Cleaning & Feature Engineering · Recap
›
Capstone: end-to-end ML cjevovod
quiz
47 / 104
🇭🇷
HR
▼
↗
Podijeli
⋯
Više
+200 XP
📋
Zadatak
📖
Teorija
🤖
AI Pomoć
Zadatak
🌐 shown in EN
📝 **Question:** Which ORDER of steps is the production-safe ML workflow? 📋 Pick the right answer. 💡 **Hint:** Re-read the theory above if unsure.
🎯 Kviz
Pitanje
📝 **Question:** Which ORDER of steps is the production-safe ML workflow? 📋 Pick the right answer. 💡 **Hint:** Re-read the theory above if unsure.
A
`model.fit(X_raw, y)` first to get a working baseline, then clean and engineer features later as accuracy improvements
B
Load → QA → clean → features → stratified split → Pipeline → cross-validate → touch test set once → persist artifact (prevents leakage)
C
Skip cross-validation, trust a single `train_test_split(random_state=0)` and ship whichever fit reaches the target metric first
D
Tune hyperparameters by reading the test-set F1 after every iteration and stop when the test metric stops improving notably
Odgovori
💬
Kontaktiraj podršku