Data preparation for MBA

The time series data was transformed to ensure that each row of data represented the customer's purchase basket. A sample of 500 active customers was selected for the analysis. The definition of active stipulated that the customer had used at least one product or service in the last 12 months and had at least one open product at the time of selection. Only customers with multiple products/services were chosen for the analysis.

The following is a snapshot of the customer data used for modeling:

Custid	Pr1	Pr2	Pr3	Pr4	Pr5
1	Residential_Mortgage	Insurance	Personal_Current_Account	Savings_Account	Premium_Current_Account
2	Personal_Current_Account	Credit_Card	Premium_Current_Account
3	Locker	Residential_Mortgage	Insurance	Credit_Card	Currency_Services
4	Personal_Current_Account	Trading_Account...