2. Apply a two-step approach:

a. Using matching distance to compute dissimilarity between observations, employ hierarchical clustering with group average linkage to produce four clusters using the variables Female, Married, Loan, and Mortgage.

b. Based on the clusters from part (a), split the original 600 observations into four separate data sets as suggested by the four clusters from part (a). For each of these four data sets, apply k-means clustering with k52 using Age, Income, and Children as variables. Normalize the values of the input variables. This will generate a total of eight clusters. Describe these eight clusters according to their “average” characteristics.

What benefit does this two-step clustering approach have over just using hierarchical clustering on all seven variables as in part (1) or just using k-means clustering on all seven variables? What weakness does it have?

Copyright