grokking-algorithms/chapter12

K-Nearest Neighbours

Useful for classification, regression and feature extraction. By examining a data point against its K nearest neighbours we can:

• categorize into a group
• predict responses
• convert the item into a list of features

A good starting point for machine learning.