Notes - Machine Learning MT23, k-nearest neighbours

Created: May 04, 2024

---

• [[Course - Machine Learning MT23]]^U

Flashcards

Suppose we have data $\langle \pmb x _ i, y _ i \rangle$ which classifies data into classes $y _ i \in \{1, \cdots, N\}$. How does $k$-nearest neighbours then classify a new point $\pmb x _ \text{new}$?

---

• Find the $k$ nearest neighbours to $\pmb x _ \text{new}$
• Output the majority among the labels for these points

How does $k$-means differ from $k$-nearest neighbours?

---

They are used for very different purposes. $k$-means is a clustering algorithm, whereas $k$-nearest neighbours is a classification algorithm.

Proofs

---

Related posts

• [[Course - Machine Learning MT23]]^U

  (outgoing)
• [[Notes - Machine Learning MT23, Classification]]^U

  (incoming)
• [[Notes - Machine Learning MT23, Spectral clustering]]^U

  (sim: 0.621)
• [[Notes - Machine Learning MT23, k-means clustering]]^U

  (sim: 0.594)
• [[Notes - Machine Learning MT23, Multidimensional scaling]]^U

  (sim: 0.516)
• [[Notes - Machine Learning MT23, Clustering]]^U

  (sim: 0.473)