validation.index: Validation Index

In fauzipandya/advfclust: Object Oriented Advanced Clustering

Description

Validation index for validating fuzzy clustering result

Usage

validation.index(object)

Arguments

object

fuzzy clustering object

Details

This function provide several validation indexs that calculated from fuzzy clustering result. Validation index can be used for choose best optimum parameter.

There are PC, MPC, CE, S, Xie Beni, Kwon, and Tang index. PC (Partition Coefficient), MPC (Modified Partition Coefficient), and CE (Classification Entropy) are calculated from membership matrix. S (Separation Index), Xie Beni, Kwon, and Tang use both distance and membership matrix.

The best cluster result can be decided with minimum value of index, except MPC and PC use maximum value.

Value

validation index object.

Slots

XB

Xie Beni Index

PC

Partition Coef.

MPC

Modifief Partition Coef.

Kwon

Kwon Index

Tang

Tang Index

S

Separation Index

CE

Classification Entropy

Author(s)

Achmad Fauzi Bagus F

References

Wang, W., & Zhang, Y. (2007). On Fuzzy Cluster Validity Indices. Fuzzy Sets and System, 2095-2117.

Examples

fuzzy.CM(iris[,1:4],K=3,m=2,max.iteration=100,threshold=1e-5,RandomNumber=1234)->cl
validation.index(cl)->valid
#example for Xie Beni index
XB(valid)

fauzipandya/advfclust documentation built on May 16, 2019, 10:56 a.m.