KNeighborsClassifier (2024)

class sklearn.neighbors.KNeighborsClassifier(n_neighbors=5, *, weights='uniform', algorithm='auto', leaf_size=30, p=2, metric='minkowski', metric_params=None, n_jobs=None)[source]#

Classifier implementing the k-nearest neighbors vote.

Read more in the User Guide.

Parameters:
n_neighborsint, default=5

Number of neighbors to use by default for kneighbors queries.

weights{‘uniform’, ‘distance’}, callable or None, default=’uniform’

Weight function used in prediction. Possible values:

  • ‘uniform’ : uniform weights. All points in each neighborhoodare weighted equally.

  • ‘distance’ : weight points by the inverse of their distance.in this case, closer neighbors of a query point will have agreater influence than neighbors which are further away.

  • [callable] : a user-defined function which accepts anarray of distances, and returns an array of the same shapecontaining the weights.

Refer to the example entitledNearest Neighbors Classificationshowing the impact of the weights parameter on the decisionboundary.

algorithm{‘auto’, ‘ball_tree’, ‘kd_tree’, ‘brute’}, default=’auto’

Algorithm used to compute the nearest neighbors:

  • ‘ball_tree’ will use BallTree

  • ‘kd_tree’ will use KDTree

  • ‘brute’ will use a brute-force search.

  • ‘auto’ will attempt to decide the most appropriate algorithmbased on the values passed to fit method.

Note: fitting on sparse input will override the setting ofthis parameter, using brute force.

leaf_sizeint, default=30

Leaf size passed to BallTree or KDTree. This can affect thespeed of the construction and query, as well as the memoryrequired to store the tree. The optimal value depends on thenature of the problem.

pfloat, default=2

Power parameter for the Minkowski metric. When p = 1, this is equivalentto using manhattan_distance (l1), and euclidean_distance (l2) for p = 2.For arbitrary p, minkowski_distance (l_p) is used. This parameter is expectedto be positive.

metricstr or callable, default=’minkowski’

Metric to use for distance computation. Default is “minkowski”, whichresults in the standard Euclidean distance when p = 2. See thedocumentation of scipy.spatial.distance andthe metrics listed indistance_metrics for valid metricvalues.

If metric is “precomputed”, X is assumed to be a distance matrix andmust be square during fit. X may be a sparse graph, in whichcase only “nonzero” elements may be considered neighbors.

If metric is a callable function, it takes two arrays representing 1Dvectors as inputs and must return one value indicating the distancebetween those vectors. This works for Scipy’s metrics, but is lessefficient than passing the metric name as a string.

metric_paramsdict, default=None

Additional keyword arguments for the metric function.

n_jobsint, default=None

The number of parallel jobs to run for neighbors search.None means 1 unless in a joblib.parallel_backend context.-1 means using all processors. See Glossaryfor more details.Doesn’t affect fit method.

Attributes:
classes_array of shape (n_classes,)

Class labels known to the classifier

effective_metric_str or callble

The distance metric used. It will be same as the metric parameteror a synonym of it, e.g. ‘euclidean’ if the metric parameter set to‘minkowski’ and p parameter set to 2.

effective_metric_params_dict

Additional keyword arguments for the metric function. For most metricswill be same with metric_params parameter, but may also contain thep parameter value if the effective_metric_ attribute is set to‘minkowski’.

n_features_in_int

Number of features seen during fit.

Added in version 0.24.

feature_names_in_ndarray of shape (n_features_in_,)

Names of features seen during fit. Defined only when Xhas feature names that are all strings.

Added in version 1.0.

n_samples_fit_int

Number of samples in the fitted data.

outputs_2d_bool

False when y’s shape is (n_samples, ) or (n_samples, 1) during fitotherwise True.

See also

RadiusNeighborsClassifier

Classifier based on neighbors within a fixed radius.

KNeighborsRegressor

Regression based on k-nearest neighbors.

RadiusNeighborsRegressor

Regression based on neighbors within a fixed radius.

NearestNeighbors

Unsupervised learner for implementing neighbor searches.

Notes

See Nearest Neighbors in the online documentationfor a discussion of the choice of algorithm and leaf_size.

Warning

Regarding the Nearest Neighbors algorithms, if it is found that twoneighbors, neighbor k+1 and k, have identical distancesbut different labels, the results will depend on the ordering of thetraining data.

https://en.wikipedia.org/wiki/K-nearest_neighbor_algorithm

Examples

>>> X = [[0], [1], [2], [3]]>>> y = [0, 0, 1, 1]>>> from sklearn.neighbors import KNeighborsClassifier>>> neigh = KNeighborsClassifier(n_neighbors=3)>>> neigh.fit(X, y)KNeighborsClassifier(...)>>> print(neigh.predict([[1.1]]))[0]>>> print(neigh.predict_proba([[0.9]]))[[0.666... 0.333...]]
fit(X, y)[source]#

Fit the k-nearest neighbors classifier from the training dataset.

Parameters:
X{array-like, sparse matrix} of shape (n_samples, n_features) or (n_samples, n_samples) if metric=’precomputed’

Training data.

y{array-like, sparse matrix} of shape (n_samples,) or (n_samples, n_outputs)

Target values.

Returns:
selfKNeighborsClassifier

The fitted k-nearest neighbors classifier.

get_metadata_routing()[source]#

Get metadata routing of this object.

Please check User Guide on how the routingmechanism works.

Returns:
routingMetadataRequest

A MetadataRequest encapsulatingrouting information.

get_params(deep=True)[source]#

Get parameters for this estimator.

Parameters:
deepbool, default=True

If True, will return the parameters for this estimator andcontained subobjects that are estimators.

Returns:
paramsdict

Parameter names mapped to their values.

kneighbors(X=None, n_neighbors=None, return_distance=True)[source]#

Find the K-neighbors of a point.

Returns indices of and distances to the neighbors of each point.

Parameters:
X{array-like, sparse matrix}, shape (n_queries, n_features), or (n_queries, n_indexed) if metric == ‘precomputed’, default=None

The query point or points.If not provided, neighbors of each indexed point are returned.In this case, the query point is not considered its own neighbor.

n_neighborsint, default=None

Number of neighbors required for each sample. The default is thevalue passed to the constructor.

return_distancebool, default=True

Whether or not to return the distances.

Returns:
neigh_distndarray of shape (n_queries, n_neighbors)

Array representing the lengths to points, only present ifreturn_distance=True.

neigh_indndarray of shape (n_queries, n_neighbors)

Indices of the nearest points in the population matrix.

Examples

In the following example, we construct a NearestNeighborsclass from an array representing our data set and ask who’sthe closest point to [1,1,1]

>>> samples = [[0., 0., 0.], [0., .5, 0.], [1., 1., .5]]>>> from sklearn.neighbors import NearestNeighbors>>> neigh = NearestNeighbors(n_neighbors=1)>>> neigh.fit(samples)NearestNeighbors(n_neighbors=1)>>> print(neigh.kneighbors([[1., 1., 1.]]))(array([[0.5]]), array([[2]]))

As you can see, it returns [[0.5]], and [[2]], which means that theelement is at distance 0.5 and is the third element of samples(indexes start at 0). You can also query for multiple points:

>>> X = [[0., 1., 0.], [1., 0., 1.]]>>> neigh.kneighbors(X, return_distance=False)array([[1], [2]]...)
kneighbors_graph(X=None, n_neighbors=None, mode='connectivity')[source]#

Compute the (weighted) graph of k-Neighbors for points in X.

Parameters:
X{array-like, sparse matrix} of shape (n_queries, n_features), or (n_queries, n_indexed) if metric == ‘precomputed’, default=None

The query point or points.If not provided, neighbors of each indexed point are returned.In this case, the query point is not considered its own neighbor.For metric='precomputed' the shape should be(n_queries, n_indexed). Otherwise the shape should be(n_queries, n_features).

n_neighborsint, default=None

Number of neighbors for each sample. The default is the valuepassed to the constructor.

mode{‘connectivity’, ‘distance’}, default=’connectivity’

Type of returned matrix: ‘connectivity’ will return theconnectivity matrix with ones and zeros, in ‘distance’ theedges are distances between points, type of distancedepends on the selected metric parameter inNearestNeighbors class.

Returns:
Asparse-matrix of shape (n_queries, n_samples_fit)

n_samples_fit is the number of samples in the fitted data.A[i, j] gives the weight of the edge connecting i to j.The matrix is of CSR format.

See also

NearestNeighbors.radius_neighbors_graph

Compute the (weighted) graph of Neighbors for points in X.

Examples

>>> X = [[0], [3], [1]]>>> from sklearn.neighbors import NearestNeighbors>>> neigh = NearestNeighbors(n_neighbors=2)>>> neigh.fit(X)NearestNeighbors(n_neighbors=2)>>> A = neigh.kneighbors_graph(X)>>> A.toarray()array([[1., 0., 1.], [0., 1., 1.], [1., 0., 1.]])
predict(X)[source]#

Predict the class labels for the provided data.

Parameters:
X{array-like, sparse matrix} of shape (n_queries, n_features), or (n_queries, n_indexed) if metric == ‘precomputed’

Test samples.

Returns:
yndarray of shape (n_queries,) or (n_queries, n_outputs)

Class labels for each data sample.

predict_proba(X)[source]#

Return probability estimates for the test data X.

Parameters:
X{array-like, sparse matrix} of shape (n_queries, n_features), or (n_queries, n_indexed) if metric == ‘precomputed’

Test samples.

Returns:
pndarray of shape (n_queries, n_classes), or a list of n_outputs of such arrays if n_outputs > 1.

The class probabilities of the input samples. Classes are orderedby lexicographic order.

score(X, y, sample_weight=None)[source]#

Return the mean accuracy on the given test data and labels.

In multi-label classification, this is the subset accuracywhich is a harsh metric since you require for each sample thateach label set be correctly predicted.

Parameters:
Xarray-like of shape (n_samples, n_features)

Test samples.

yarray-like of shape (n_samples,) or (n_samples, n_outputs)

True labels for X.

sample_weightarray-like of shape (n_samples,), default=None

Sample weights.

Returns:
scorefloat

Mean accuracy of self.predict(X) w.r.t. y.

set_params(**params)[source]#

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects(such as Pipeline). The latter haveparameters of the form <component>__<parameter> so that it’spossible to update each component of a nested object.

Parameters:
**paramsdict

Estimator parameters.

Returns:
selfestimator instance

Estimator instance.

set_score_request(*, sample_weight: bool | None | str = '$UNCHANGED$') KNeighborsClassifier[source]#

Request metadata passed to the score method.

Note that this method is only relevant ifenable_metadata_routing=True (see sklearn.set_config).Please see User Guide on how the routingmechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains theexisting request. This allows you to change the request for someparameters and not others.

Added in version 1.3.

Note

This method is only relevant if this estimator is used as asub-estimator of a meta-estimator, e.g. used inside aPipeline. Otherwise it has no effect.

Parameters:
sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns:
selfobject

The updated object.

Gallery examples#

KNeighborsClassifier (1)

Release Highlights for scikit-learn 0.24

Release Highlights for scikit-learn 0.24

KNeighborsClassifier (2)

Classifier comparison

Classifier comparison

KNeighborsClassifier (3)

Plot the decision boundaries of a VotingClassifier

Plot the decision boundaries of a VotingClassifier

KNeighborsClassifier (4)

Caching nearest neighbors

Caching nearest neighbors

KNeighborsClassifier (5)

Comparing Nearest Neighbors with and without Neighborhood Components Analysis

Comparing Nearest Neighbors with and without Neighborhood Components Analysis

KNeighborsClassifier (6)

Dimensionality Reduction with Neighborhood Components Analysis

Dimensionality Reduction with Neighborhood Components Analysis

KNeighborsClassifier (7)

Nearest Neighbors Classification

Nearest Neighbors Classification

KNeighborsClassifier (8)

Importance of Feature Scaling

Importance of Feature Scaling

KNeighborsClassifier (9)

Digits Classification Exercise

Digits Classification Exercise

KNeighborsClassifier (10)

Classification of text documents using sparse features

Classification of text documents using sparse features

KNeighborsClassifier (2024)

FAQs

How to determine the optimal value for k in k nearest neighbor algorithm? ›

The optimal K value usually found is the square root of N, where N is the total number of samples. Use an error plot or accuracy plot to find the most favorable K value. KNN performs well with multi-label classes, but you must be aware of the outliers.

How to improve KNN results? ›

What are the most effective ways to improve k-nearest neighbor search accuracy?
  1. Choose the right k value.
  2. Use a suitable distance metric.
  3. Scale and normalize the data. Be the first to add your personal experience.
  4. Reduce the dimensionality. ...
  5. Use an efficient data structure. ...
  6. Use an ensemble method. ...
  7. Here's what else to consider.
Dec 28, 2023

What are the drawbacks of the K nearest neighbor algorithm? ›

KNN has some drawbacks and challenges, such as computational expense, slow speed, memory and storage issues for large datasets, sensitivity to the choice of k and the distance metric, and susceptibility to the curse of dimensionality.

What does KNeighborsClassifier do in Python? ›

KNeighborsClassifier is a supervised learning algorithm that makes classifications based on data neighbors.

How can you avoid overfitting in KNN? ›

To prevent overfitting, we can smooth the decision boundary by K nearest neighbors instead of 1. Find the K training samples , r = 1 , … , K closest in distance to , and then classify using majority vote among the k neighbors.

What is the optimal value of K in KNN cross-validation? ›

There is no single formula to determine the best value of K. The value of K that works best will depend on the specific problem being solved and the characteristics of the data being analyzed, in this research paper author choose Cross-Validation and the result shown in table 1.

Why is my KNN accuracy so low? ›

The relatively low accuracy of kNN is caused by several factors. One of them is that every characteristic of the method has the same result on calculating distance. The solution of this problem is to give weight to each data characteristic [12].

What is a good accuracy rate for KNN? ›

Many researchers have proposed enhancements to KNN algorithm to explain how the decision criteria can be modified. Any classification accuracy above 80% is acceptable. It meant to say your feature extraction process from the data is good and you have chosen optimal feature required.

What happens when you increase the value of K in KNN? ›

The way KNN works is by taking votes from adjacent points. If your data is well separated then as you increase K then the model won't be as confused. In some cases, the performance would decrease; this is when K is so big that it's more than the number of samples inside a class.

When should you not use KNN? ›

One such situation is when dealing with large datasets and high-dimensional data, as KNN becomes computationally expensive and less effective in these cases 3. Another situation is when the classes in the dataset are highly unbalanced, with one class having significantly fewer examples than the others.

What is the problem with KNN? ›

Disadvantages of the KNN Algorithm

Does not scale – As we have heard about this that the KNN algorithm is also considered a Lazy Algorithm. The main significance of this term is that this takes lots of computing power as well as data storage.

Why is KNN called lazy learner? ›

K-NN is a non-parametric algorithm, which means that it does not make any assumptions about the underlying data. It is also called a lazy learner algorithm because it does not learn from the training set immediately instead it stores the data set and at the time of classification it performs an action on the data set.

How to improve KNN performance? ›

  1. 1 Choose the right k. The number of neighbors, or k, is a crucial parameter for KNN. ...
  2. 2 Preprocess the data. ...
  3. 3 Reduce the dimensionality. ...
  4. 4 Use an index structure. ...
  5. 5 Here's what else to consider.
Apr 18, 2023

How to decide k value in KNN? ›

You can use the common formula k = sqrt(n) where n is the number of data points in your training set or you can try choosing k where there is a good balance between computation expense vs noise. Consider your what fits your problem: Do you care about runtime?

Does KNN need normalization? ›

When training a kNN classifier, it's essential to normalize the features. This is because kNN measures the distance between points. The default is to use the Euclidean Distance, which is the square root of the sum of the squared differences between two points.

How to find optimal k value for K-means algorithm? ›

This method is a visual technique used to determine the best K value for a k-means clustering algorithm. In this method, a graph known as the elbow graph plots the within-cluster-sum-of-square (WCSS) values against various K values. The optimal K value is identified at the point where the graph bends like an elbow.

What is the default value of K in KNN? ›

Note that if k is chosen as the total number of observations in the Training Set, all the observations in the Training Set become nearest neighbors. The default value for this option is 1.

What is K nearest neighbor optimization? ›

K-Nearest Neighbors algorithm (KNN) is one of the simplest algorithms; it is widely used in predictive analysis. To optimize its performance and to accelerate its process, this paper proposes a new solution to speed up KNN algorithm based on clustering and attributes filtering.

Top Articles
Latest Posts
Recommended Articles
Article information

Author: Rob Wisoky

Last Updated:

Views: 5241

Rating: 4.8 / 5 (48 voted)

Reviews: 87% of readers found this page helpful

Author information

Name: Rob Wisoky

Birthday: 1994-09-30

Address: 5789 Michel Vista, West Domenic, OR 80464-9452

Phone: +97313824072371

Job: Education Orchestrator

Hobby: Lockpicking, Crocheting, Baton twirling, Video gaming, Jogging, Whittling, Model building

Introduction: My name is Rob Wisoky, I am a smiling, helpful, encouraging, zealous, energetic, faithful, fantastic person who loves writing and wants to share my knowledge and understanding with you.