In this example data, there are six customers, and each is described with four attributes:
Customer, Age, Income, Cards, Gender. Your task is to do clustering of the 6 customers.
Follow the steps to carry out clustering and use as much as details to describe the process.
Customer Gender Age Income(1000) Cards
cusA
0 40
50
3
cusB
0 38
48
2
cusC
1 25
30
4
cusD
0 69
170
1
cusE
0 19
40
2
cusF
0 23
140
3
Manhattan distance metrics is adopted in the clustering. $d_{Manhattan}(A, B) = \sum_{i=1}^{k} |x_i - y_i|$
Bottom-up approach Algorithm is adopted in the clustering.
Trial-and-error: K=2.
Initial Step:
Number of clusters:
cusA cusB cusC cusD cusE cusF
cusA
cusB
cusC
cusD
cusE
cusF
Table 1. The distance matrix contains the distance of each pair of clusters
1. Step 1
New cluster
distance in Table 1.
Number of clusters:
formed by merging two clusters which have the smallest
Update the distance matrix, which is denoted as Table 2.
2. Step 2
New cluster
distance in Table 2.
Number of clusters:
formed by merging two clusters which have the smallest
Update the distance matrix, which is denoted as Table 3.
3. Step 3
New cluster
distance in Table 3.
Number of clusters:
formed by merging two clusters which have the smallest
Update the distance matrix, which is denoted as Table 4.
4. Step 4
New cluster
distance in Table 4.
Number of clusters:
formed by merging two clusters which have the smallest
Update the distance matrix, which is denoted as Table 5.
5. Step 5
New cluster
formed by merging two clusters which have the smallest
distance in Table 5.
Number of clusters:
(Stop condition is met)
