Artificial Intelligence By Example

The KNN algorithm

Generally, a word requires a context to mean something. Looking for "neighbors" close by provides an efficient way to determine where the word belongs.

KNN is supervised because it uses the labels of the data provided to train its algorithm. KNN, in this case, is used for classification purposes. For a given point p, KNN will calculate the distances to all other points. Then k represents the k-nearest neighbors to take into account.

Let's clear this up through an example. In English, the word "coach" can mean a trainer on a football field, a bus, or a railroad passenger car. Since the prototype is for a transportation company, "coach" will mostly be a bus that should not be confused with a trainer:

• Step 1: Parsing (examining ...