K-Nearest Neighbours and K-Means Clustering

Contents

4.1. K-Nearest Neighbours and K-Means Clustering#

In this lesson we’ll be looking at more machine learning techniques:

K-nearest neighbours regressor - supervised learning
K-nearest neighbours classifier - supervised learning
K-means clustering - unsupervised learning

The core concept underpinning these models is the idea that the distance between datapoints is a measure of similarity.

../../_images/similarity.png

4.1.1. Glossary#

K-nearest neighbours#: A machine learning algorithm that makes predictions using similar samples nearby in the training data.
Distance#: A measure used to compare how close or similar samples are.
Similarity#: How alike samples are, often measured using distance.
K-means clustering#: An unsupervised learning algorithm that automatically groups data using the distance between samples to measure similarity.