EasyDeepLearn

Why does KNN degrade badly in high dimensions?

medium

Answer

  • In high dimensions, pairwise distances between random points concentrate — the ratio (max - min) / min distance approaches 0.
  • Every point becomes roughly the same distance from every other, so the concept of 'nearest' loses meaning and KNN votes become dominated by noise.
  • Remedies: apply dimensionality reduction (PCA, autoencoder) first, use domain-specific distance metrics, or switch to a model with stronger inductive bias (tree ensembles).
Check yourself — multiple choice
  • Distances become more discriminative in high dimensions
  • Pairwise distances concentrate — the nearest neighbour is not much closer than the farthest
  • KNN can only run on GPUs
  • It just needs more training data

Distance concentration in high dimensions kills the notion of a meaningful 'nearest' neighbour.

#knn#theory

Practise Supervised Learning

215 interview questions in this topic.

Related questions