Statements (52)
Predicate | Object |
---|---|
gptkbp:instanceOf |
gptkb:algorithm
unsupervised learning algorithm |
gptkbp:application |
image compression
anomaly detection customer segmentation document clustering market basket analysis |
gptkbp:category |
centroid-based clustering
partitioning method |
gptkbp:complexity |
O(n*k*i*d)
|
gptkbp:convergenceCriterion |
centroids do not move
no change in assignments |
https://www.w3.org/2000/01/rdf-schema#label |
K-means clustering
|
gptkbp:implementedIn |
gptkb:MATLAB
gptkb:Spark_MLlib gptkb:scikit-learn R |
gptkbp:input |
set of data points
number of clusters (k) |
gptkbp:introduced |
gptkb:Stuart_Lloyd
|
gptkbp:introducedIn |
1957
|
gptkbp:limitation |
sensitive to outliers
assumes spherical clusters not suitable for non-globular clusters requires number of clusters as input sensitive to initial centroids |
gptkbp:measures |
Euclidean distance
|
gptkbp:objective |
minimize within-cluster sum of squares
|
gptkbp:output |
cluster centroids
k clusters |
gptkbp:popularizedBy |
gptkb:J._MacQueen
1967 |
gptkbp:relatedTo |
gptkb:Gaussian_Mixture_Model
gptkb:Expectation-Maximization_algorithm Hierarchical clustering |
gptkbp:step |
assign points to nearest centroid
initialize centroids repeat until convergence update centroids |
gptkbp:supportsAlgorithm |
iterative algorithm
|
gptkbp:usedIn |
gptkb:machine_learning
pattern recognition data mining image segmentation vector quantization |
gptkbp:variant |
gptkb:Bisecting_K-means
gptkb:Fuzzy_C-means gptkb:K-medoids gptkb:Mini-batch_K-means |
gptkbp:bfsParent |
gptkb:H2O-3
gptkb:Gaussian_mixture_models |
gptkbp:bfsLayer |
6
|