KNN (K-Nearest Neighbors)

GPTKB entity

Statements (43)
Predicate Object
gptkbp:instanceOf gptkb:model
gptkbp:advantage simple to implement
no training phase
sensitive to irrelevant features
sensitive to the scale of data
computationally expensive at prediction time
gptkbp:alternativeName gptkb:k-NN
k nearest neighbors
gptkbp:category instance-based learning
lazy learning
gptkbp:decision majority vote
average of neighbors (for regression)
https://www.w3.org/2000/01/rdf-schema#label KNN (K-Nearest Neighbors)
gptkbp:implementedIn gptkb:MATLAB
gptkb:scikit-learn
R
gptkbp:input feature vectors
gptkbp:introduced Fix and Hodges
gptkbp:introducedIn 1951
gptkbp:measures gptkb:Hamming_distance
gptkb:Manhattan_distance
gptkb:Minkowski_distance
Euclidean distance
gptkbp:notableFor pattern recognition
recommendation systems
data mining
image classification
text categorization
gptkbp:output class label
regression value
gptkbp:parameter number of neighbors (k)
gptkbp:preprocessing normalization
feature scaling
gptkbp:relatedTo gptkb:tree
support vector machine
nearest centroid classifier
gptkbp:requires labeled data
gptkbp:sensitivity gptkb:curse_of_dimensionality
gptkbp:type supervised learning
gptkbp:usedFor gptkb:dictionary
regression
gptkbp:bfsParent gptkb:HNSW
gptkbp:bfsLayer 7