Mastering Machine Learning with scikit-learn - Second Edition

Classification and Regression with k-Nearest Neighbors

In this chapter, we will introduce k-Nearest Neighbors (KNN), a simple algorithm that can be used for classification and regression tasks. KNN's simplicity belies its power and usefulness; it is widely used in the real world in a variety of applications, including search and recommender systems. We will compare and contrast KNN with simple linear regression and work through toy problems to understand the model.