package prbnmcn-clustering

  1. Overview
  2. Docs

Parameters

module E : Intf.Metric

Signature

val k_medoids : precompute:bool -> elements:E.t array -> k:int -> init:init -> algorithm:algorithm -> termination:termination -> Stdlib.Random.State.t -> E.t array array

k_means performs the clustering using the provided initialization method.

val cost : classes:E.t array array -> float

cost returns the sum over all classes of the sum of distances from the medoid of the class to all elements of the class.