KNN & Decision Trees Quiz
Check your understanding of instance-based and tree-based learning — the bias–variance effect of k, why scaling matters for distances, and how trees choose their splits.
Exercise
What happens to a KNN classifier as k decreases to 1?
Need a hint?
With k = 1, the training error is exactly zero. Suspicious?
Exercise
Why does KNN usually require feature scaling?
Need a hint?
Imagine one feature in meters (0–2) and another in dollars (0–100,000).
Exercise
Information gain measures what about a candidate split?
Need a hint?
Entropy before vs. entropy after.
Exercise
What is the Gini impurity of a node containing only one class?
Need a hint?
Gini = 1 − Σpᵢ². What if one pᵢ = 1?
Exercise
An unconstrained decision tree grown to purity usually does what?
Need a hint?
Each leaf may end up holding a single training point.
Exercise
Increasing k in KNN (or capping a decision tree's depth) tunes model complexity. Which slot of the ML project loop does this dial live in?
Need a hint?
You are changing how expressive the model family is, not the loss or the data.
Exercise
Transfer test — a technique this course did NOT teach: focal loss down-weights easy examples so a detector focuses on hard, rare ones. Which slot does it modify?
Need a hint?
It reshapes the loss to care about different errors.