Lesson 16 — k-NN (tiny)

k-nearest neighbours: ask your neighbours

k-NN classifies a new point by looking at the k closest labelled points and taking a majority vote of their labels. No training needed — the data itself is the model. sorted(points, key=...) orders points by a rule you give it; here, by distance to the new point.

from collections import Counter
Counter(["a", "b", "a"]).most_common(1)[0][0]   # "a"

Why it matters: k-NN is the simplest real classifier and the idea behind "find similar items" features everywhere.