Feature Engineering & Scaling
Machine Learning · Intermediate
Definition
Feature engineering creates and transforms inputs so models learn patterns better.
Good features often matter more than the choice of model.
Diagram
Flowchart
Example: Min-max scaling
x = [10, 20, 30]
lo, hi = min(x), max(x)
print([(v - lo) / (hi - lo) for v in x])Output: [0.0, 0.5, 1.0]
Key points
- Fit scalers on training data only
- One-hot encode nominal categories
Interview questions
Standardization formula?
z = (x − mean) / std.