@Title("K-means Initialization Strategies")
Package elki.clustering.kmeans.initialization
Initialization strategies for k-means.
-
Interface Summary Interface Description KMeansInitialization Interface for initializing K-Means -
Class Summary Class Description AbstractKMeansInitialization Abstract base class for common k-means initializations.AbstractKMeansInitialization.Par Parameterization class.AFKMC2 AFK-MC² initializationAFKMC2.Instance Abstract instance implementing the weight handling.AFKMC2.Par Parameterization class.FarthestPoints<O> K-Means initialization by repeatedly choosing the farthest point (by the minimum distance to earlier points).FarthestPoints.Par<O> Parameterization class.FarthestSumPoints<O> K-Means initialization by repeatedly choosing the farthest point (by the sum of distances to previous objects).FarthestSumPoints.Par<V> Parameterization class.FirstK<O> Initialize K-means by using the first k objects as initial means.FirstK.Par<V extends NumberVector> Parameterization class.KMC2 K-MC² initializationKMC2.Instance Abstract instance implementing the weight handling.KMC2.Par Parameterization class.KMeansPlusPlus<O> K-Means++ initialization for k-means.KMeansPlusPlus.Instance<T> Abstract instance implementing the weight handling.KMeansPlusPlus.MedoidsInstance Instance for k-medoids.KMeansPlusPlus.NumberVectorInstance Instance for k-means, number vector based.KMeansPlusPlus.Par<V> Parameterization class.Ostrovsky Ostrovsky initial means, a variant of k-means++ that is expected to give slightly better results on average, but only works for k-means and not for, e.g., PAM (k-medoids).Ostrovsky.Par Parameterization class.Predefined Run k-means with prespecified initial means.Predefined.Par Parameterization class.RandomlyChosen<O> Initialize K-means by randomly choosing k existing elements as initial cluster centers.RandomlyChosen.Par<V> Parameterization class.RandomNormalGenerated Initialize k-means by generating random vectors (normal distributed with \(N(\mu,\sigma)\) in each dimension).RandomNormalGenerated.Par Parameterization class.RandomUniformGenerated Initialize k-means by generating random vectors (uniform, within the value range of the data set).RandomUniformGenerated.Par Parameterization class.SampleKMeans<V extends NumberVector> Initialize k-means by running k-means on a sample of the data set only.SphericalAFKMC2 Spherical K-Means++ initialization with markov chains.SphericalAFKMC2.Instance Abstract instance implementing the weight handling.SphericalAFKMC2.Par Parameterization class.SphericalKMeansPlusPlus<O> Spherical K-Means++ initialization for k-means.SphericalKMeansPlusPlus.Instance Abstract instance implementing the weight handling.SphericalKMeansPlusPlus.Par<V> Parameterization class.