Computer Science From Scratch

7 CLASSIFICATION WITH K-NEAREST NEIGHBORS

This chapter will introduce k-nearest neighbors (KNN), an extraordinarily simple machine learning algorithm that can be highly effective for some applications. First developed in the 1950s and 1960s,¹ KNN can be used for both classification (deciding what category something belongs to) and regression (predicting a value). For readers daunted by the complexities of machine learning, KNN provides an accessible yet real-world entry point into the field. In this chapter, we’ll use KNN to solve two classification problems with a high degree of accuracy: distinguishing between different types of fish ...