Building & Improving a K-Nearest Neighbors Algorithm in Python

The K-Nearest Neighbors algorithm, K-NN for short, is a classic machine learning work horse algorithm that is often overlooked in the day of deep learning. In this tutorial, we will build a K-NN algorithm in Scikit-Learn and run it on the MNIST dataset. From there, we will build our own K-NN algorithm in the hope of developing a classifier with both better accuracy and classification speed than the Scikit-Learn K-NN. The K-Nearest Neighbors algorithm is a supervised machine learning algorithm that is simple to implement, and yet has the ability to make robust classifications. One of the biggest advantages of K-NN is that it is a lazy-learner.