BitWithBite
AI & Machine Learning Worksheet
🤖 Preprocessing: Scaling & Encoding
SCORE___ / 20
NameClassDate
After this worksheet you can
StandardScaler and MinMaxWhy scaling mattersOne-hot encodingHandle unseen categories📚 Quick Recap
Raw data almost never goes directly into a model. Scaling stops one feature dominating because it has bigger numbers. Encoding lets models understand categories. Done wrong — especially fitting on test data — it corrupts every result you produce. Done right, it's invisible infrastructure that makes every model work better.
🧠 Section A · Concept Check ● BEGINNER 4 × 1 = 4
1StandardScaler outputs features with:
2MinMaxScaler maps to:
3Categorical text becomes numbers via:
4Distance-based models (KNN/SVM) need:
🧮 Section B · Problem Solving ● INTERMEDIATE 2 + 3×3 = 11
5Standardisation: (x - mean) / .
6One-hot creates one column per category.
7Why does KNN fail with unscaled salary + age?
8One-hot encode colors red/green/blue - result shape?
9Standard vs MinMax - when which?
🚀 Section C · Challenge ● CHALLENGE 5
10What is data leakage in scaling?
💭 Reflection — the most useful thing I learned:
A ___/4 B ___/11 C ___/5 Total ___/20
Teacher's Signature
Parent's Signature
✂ answer key — fold or cut before handing out
1-A 2-A 3-A 4-A | 5 std 6 binary/indicator 7 = Salary's huge range dominates the distance metric 8 = 3 binary columns 9 = Standard for normal-ish data/linear models; MinMax for bounded inputs | 10 = Fitting the scaler with test data - inflates evaluation