Source

K Nearest Neighbors#

A distance-based algorithm that locates the k nearest neighbors (data points) from the training set and uses a weighted vote to classify unknown samples during inference. A kernelized distance function allows the user to specify different concepts of distance to the estimator.

Note: This learner is considered a lazy learner because it does the majority of its computation during inference. For a fast spatial tree-based version, see KD Neighbors.

Interfaces: Estimator, Learner, Online, Probabilistic, Persistable

Data Type Compatibility: Continuous

Parameters#

# Param Default Type Description
1 k 5 int The number of nearest neighbors to consider when making a prediction.
2 weighted true bool Should we use the inverse distances as confidence scores when making predictions?
3 kernel Euclidean object The distance kernel used to compute the distance between sample points.

Additional Methods#

This estimator does not have any additional methods.

Example#

use Rubix\ML\Classifiers\KNearestNeighbors;
use Rubix\ML\Kernels\Distance\Manhattan;

$estimator = new KNearestNeighbors(3, true, new Manhattan());