Why Scale Features? — Interview Answer — ML Technical Interview Questions & Answers

Model answer covering distance-based algorithms.

Why Scale Features? — Interview Answer

  • Say this: “Scaling puts features on a similar range so distance and gradient methods don’t treat fees 80000 as more important than hours 8. I fit the scaler on train and transform test.”
  • Need it: KNN, SVM, logistic (helps), PCA, gradient descent nets. Trees: usually optional (splits don’t care about units).
  • StandardScaler (mean 0, std 1) or MinMax 0–1. Pipeline it.
  • Trap: scaling with test statistics. Trap: scaling dummy 0/1 and tree-only pipelines as if it were mandatory without explanation.

Technical Interview Tip: KNN/SVM/PCA: scale. Fit on train. Trees can skip.

Code Examples for Why Scale Features? — Interview Answer

Why Scale Features? — Interview Answer — working example

xs = [10, 20, 30, 40]
mean = sum(xs) / len(xs)
var = sum((x - mean) ** 2 for x in xs) / len(xs)
std = var ** 0.5
z = [(x - mean) / std for x in xs]
lo, hi = min(xs), max(xs)
minmax = [(x - lo) / (hi - lo) for x in xs]
print([round(v, 3) for v in z])
print(minmax)
# sklearn.preprocessing.StandardScaler fits mean/std on train only

Why Scale Features? — Interview Answer — scaling keeps features comparable — critical for distance-based models and many optimizers.

How to Explain Why Scale Features? — Interview Answer to an Interviewer

Why scale?

Distance and gradients Units differ

Which models need it most?

KNN SVM PCA GD nets