Question

You are presented with a data-set containing 6 observations, and wish to cluster them using hierarchal clustering. A similarity function s(x,x') is chosen which measures how similar observation r is to observation ', for each pair of observations. The distance function is then defined as d(x, x)-1- s(x,') The similarity scores are give by the below matrix, where the [ijlth entry is the similarty between observations r and a 123456 11.00 0.70 0.65 0.40 0.20 0.05 2 0.70 1.00 0.95 0.70 0.50 0.35 3 0.65 0.95 1.00 0.75 0.55 0.40 0.40 0.70 0.75 1.00 0.80 0.65 5 0.20 0.50 0.55 0.80 1.00 0.85 6 0.05 0.35 0.40 0.65 0.85 1.00 Plot the dendogram produced by hierarchal clustering using: 1. Single link clustering [Assessed: 2 marks] 2. Complete link clustering [Assessed: 2 marks]

Answer 1

By using matlab, we will find dendrogram for

a) single link clustering

b)complete link clustering

a)

b)

Dendrogram Dendrogram Single Linkage, Correlation Coefficient Distance 59.51 73.01- 86.50-1 г 100.00 C1 02 04 C5 C6 Variables

steps:

Dendrogram Complete Linkage, Correlation Coefficient Distance 0.00 33.33- 66.67 100.00 C1 DB C4 C5 C6 Variables

1) open a minitab sheet. Enter a data in sheet

2) check the upper ribbon of sheet. There will be a option of stat open it.

3) Then there will be a option as Multivariate ​​​​​​, under multivariate cluster variable option will be seen. Click on that.

4) a new table will be opened, give the collumn name or select the collumn name in that table.select linkage method & click on show dendrogram.

It will give required output.