K Nearest Neighbors

Questions

Notes

This is a learning algorithm that allows you to classify points. When there is an unlabeled point in the space the k-Nearest Neighbors take a "vote" to put the point into their class and the class with the most votes wins.

Applications

In data preprocessing where there are missing data points you can use the k-nearest neighbors algorithm to estimate those points.

Advantages/Disadvantages

Easy to implement, no training required.

Overfits easily, curse of high dimensionality,

See Also