K-Nearest Neighbors Classifier

The K-Nearest Neighbors Classifier is an algorithm that uses similarity to make predictions about the grouping of unknown data. In other words, based upon the characteristics of a certain number (K) of known elements that are the most close-by (Nearest Neighbor) to new unknown data, it predicts the category of it (Classifier). The poster guides the reader in understanding how it works as if they were starting out a new career in a video game in which it’s necessary to guess whether unknown ships are enemies or allies. By following the steps given by the crew captain, they are thus given a thorough explanation of the prerequisites, the calculation and the process of verifying and optimising its accuracy, along with a couple of insider tips.