If you were to present the results to your internal client, how would you label Clusters 1 and 2?

In your opinion, what should you do about Cluster 3 containing only one observation?

What business recommendations can you make based on the results?

Demographic:

education

customer_id

age

marital

job

Other information:

Default default? (yeso)

Balance checking account balance

housing having a mortgage account (yeso)

loan having a loan account (yeso)

Promotion:

Contact the channel salesperson contacted the customer

Day contact date

Month contact month

Duration number of days since the last contact

Campaign number of campaigns

pdays see the detail below

previous number of previous campaigns

poutcome the previous outcome

y currently having an active promotion of some sort (yeso)

pdays

number of days that passed by after the client was last contacted from a previous campaign (numeric, -1 means client was not previously contacted). This is a combination of a continuous variable and a categorical variable. He consulted with the domain expert on the team and she suggested that the variable be binned as follows:

E Profile: _SEGMENT_ 507 80- 40- 20 - 60 - 30 Segment 1 Count: 7402 Percent: 69.98 Percent 40 - 20 20- 10 0 0 OUR Transformed previous Transformed balance Age pdays poutcome JOB month 80+ 20 30- 15 - 60 - Segment: 2 20 Count: 3175 Percent: 30.02 2 Percent Percent 10 - Percent 40 10 - 20 - 0 0 Transformed balance Age 0 Transformed previous pdays poutcome JOB month Segment: 3 Count: 1 Percent: 0.01 Q y contact martial y contact martial E Profile: _SEGMENT_ 507 80- 40- 20 - 60 - 30 Segment 1 Count: 7402 Percent: 69.98 Percent 40 - 20 20- 10 0 0 OUR Transformed previous Transformed balance Age pdays poutcome JOB month 80+ 20 30- 15 - 60 - Segment: 2 20 Count: 3175 Percent: 30.02 2 Percent Percent 10 - Percent 40 10 - 20 - 0 0 Transformed balance Age 0 Transformed previous pdays poutcome JOB month Segment: 3 Count: 1 Percent: 0.01 Q y contact martial y contact martial