Nearest Neighbors

k-nearest neighbor classification

To train a k-nearest neighbor model, use the Classification Learner app. For greater flexibility, train a k-nearest neighbor model using fitcknn in the command-line interface. After training, predict labels or estimate posterior probabilities by passing the model and predictor data to predict.

Apps

Classification Learner

Train models to classify data using supervised machine learning

Blocks

ClassificationKNN Predict

Classify observations using nearest neighbor classification model (Since R2022b)

Functions

expand all

Create Nearest Neighbor Model

fitcknn Fit k-nearest neighbor classifier

Create Nearest Neighbor Searcher

`ExhaustiveSearcher`	Create exhaustive nearest neighbor searcher
`KDTreeSearcher`	Create Kd-tree nearest neighbor searcher
`hnswSearcher`	Hierarchical Navigable Small Worlds (HNSW) approximate nearest neighbor search (Since R2024a)
`createns`	Create nearest neighbor searcher object

Interpret Nearest Neighbor Model

`lime`	Local interpretable model-agnostic explanations (LIME) (Since R2020b)
`partialDependence`	Compute partial dependence (Since R2020b)
`permutationImportance`	Predictor importance by permutation (Since R2024a)
`plotPartialDependence`	Create partial dependence plot (PDP) and individual conditional expectation (ICE) plots
`shapley`	Shapley values (Since R2021a)

Cross-Validate

`crossval`	Cross-validate machine learning model
`kfoldEdge`	Classification edge for cross-validated classification model
`kfoldLoss`	Classification loss for cross-validated classification model
`kfoldfun`	Cross-validate function for classification
`kfoldMargin`	Classification margins for cross-validated classification model
`kfoldPredict`	Classify observations in cross-validated classification model

Measure Performance

`loss`	Loss of k-nearest neighbor classifier
`resubLoss`	Resubstitution classification loss
`compareHoldout`	Compare accuracies of two classification models using new data
`edge`	Edge of k-nearest neighbor classifier
`margin`	Margin of k-nearest neighbor classifier
`resubEdge`	Resubstitution classification edge
`resubMargin`	Resubstitution classification margin
`testckfold`	Compare accuracies of two classification models by repeated cross-validation

Classify Observations

`predict`	Predict labels using k-nearest neighbor classification model
`resubPredict`	Classify training data using trained classifier

Gather Properties of Nearest Neighbor Model

gather Gather properties of Statistics and Machine Learning Toolbox object from GPU (Since R2020b)

Related Functions

`pdist`	Pairwise distance between pairs of observations
`pdist2`	Pairwise distance between two sets of observations

Objects

expand all

Nearest Neighbor Classification Models

`ClassificationKNN`	k-nearest neighbor classification
`ClassificationPartitionedModel`	Cross-validated classification model

Topics

Train Nearest Neighbor Classifiers Using Classification Learner App
Create and compare nearest neighbor classifiers, and export trained models to make predictions for new data.
Visualize Decision Surfaces of Different Classifiers
This example shows how to visualize the decision surface for different classification algorithms.
Supervised Learning Workflow and Algorithms
Understand the steps for supervised learning and the characteristics of nonparametric classification and regression functions.
Classification Using Nearest Neighbors
Categorize data points based on their distance to points in a training data set, using a variety of distance metrics.

Related Information

Speaker Identification Using Pitch and MFCC (Audio Toolbox)