Comparing clustering methods

After fitting data into clusters using different clustering methods, you may wish to measure the accuracy of the clustering. In most cases, you can use either intra-cluster or inter-cluster metrics as measurements. We now introduce how to compare different clustering methods using custer.stat from the fpc package.

Getting ready

In order to perform the clustering method comparison, you need to have completed the previous recipe by generating the hotel location dataset.

How to do it…

Perform the following steps to compare clustering methods:

1. First, install and load the fpc package:

   > install.packages("fpc")
   > library(fpc)
   

2. You then need to use hierarchical clustering with the single method to cluster customer data and generate the object hc_single:

   > hotel.dist <- dist(hotel[,c('lat', 'lon')] , method="euclidean")
   > single_c <- hclust(hotel.dist, method="single")
   
   > hc_single <- cutree(single_c, k = 3)
   

3. Use hierarchical clustering with the complete...