Fuzzy Clustering Analysis"

In RcmdrPlugin.FuzzyClust: R Commander Plug-in for Fuzzy Clustering Methods (Fuzzy C-Means and Gustafson Kessel)

Result of Fuzzy Clustering Analysis

Cluster Description

load("cluster.Rda")
pp<-ncol(cluster.fuzzy[[1]]$Clust.desc)
k<-ncol(cluster.fuzzy[[1]]$U)
cat(paste("Function Objective\t:",cluster.fuzzy[[1]]$func.obj,"\n"))
cat(paste("Fuzzyfier\t:",cluster.fuzzy[[1]]$m,"\n"))
cat(paste("N Cluster\t:",k,"\n"))
cat("Hard Label Partition\t:\n")
Label<-cluster.fuzzy[[1]]$Clust.desc[,ncol(cluster.fuzzy[[1]]$Clust.desc)]
library(knitr)
kable(Label)

Cluster Biplot

Note: this plot can be used to interpret your cluster on data that visualize in 2D via Principal Component Analysis.

Radar Plot: Cluster Centroid

Note: Radar Plot can interpret your centroid in a Radar Plot. The value "0" on label means the mean value of variable. The value "+/- 0.5" means the mean value of variable +/- 0.5 standar deviation. And the value "+/- 1" means the mean value of variable +/- standar deviation. The radar plot made this way to be easy understanding and easy on comparison between Cluster Centroid.

Cluster Centroid

Note: This is the exact value of centroid.

library(knitr)
Cluster<-colnames(cluster.fuzzy[[1]]$U)
colnames(cluster.fuzzy[[1]]$V)->kolomvariabel
paste("V",c(1:ncol(cluster.fuzzy[[1]]$V)),sep="")->variabel
colnames(cluster.fuzzy[[1]]$V)<-variabel
kable(cbind(Cluster,round(cluster.fuzzy[[1]]$V,2)))
cat("\nKeterangan Variabel\n")
kable(cbind(kolomvariabel, variabel))

Fuzzy Partition Matrix

Partition can be interpret the value of probability of membership among cluster. The highest partition mean the more probability to grouping to that cluster.

library(knitr)
observation<-rownames(cluster.fuzzy[[1]]$Clust.desc)
kable(cbind(observation,round(cluster.fuzzy[[1]]$U,2)))

Index Validation

Note: MPC stands from Modified partition coefficient, CE stands from Classification Entropy, XB stands from Xie Beni, and S stands from Separation index.

cat("MPC Index\t:",cluster.fuzzy[[2]][1],"\n")
cat("CE Index\t:",cluster.fuzzy[[2]][2],"\n")
cat("XB Index\t:",cluster.fuzzy[[2]][3],"\n")
cat("S Index\t:",cluster.fuzzy[[2]][4],"\n")

RcmdrPlugin.FuzzyClust documentation built on May 2, 2019, 4:55 p.m.