MS5107 briefly outlining the process of model development in the context of the CRISP-DM methodology.

Question 2: Data mining and predictive analytics.

The management of Western Alliance Bank is concerned with how to optimise marketing strategies and improve its effectiveness in selling term deposits to customers. You have been asked to help the bank in developing a more granular understanding of its customer base, predict customers’ responses to its telemarketing campaign, and establish a target customer profile for future marketing plans. The bank can then focus its marketing efforts on those customers.

Dataset:
The bank provides a dataset (file Bank.xlsx) containing data from previous telemarketing campaigns (phone calls). The dataset contains input variables, such as age, job, marital, education, etc. It also contains output variable y that shows if the customer has subscribed for term deposit or not (binary: “yes”, “no”). The description of each variable is given in the ‘Description’ worksheet.

Task: Build a predictive model.
Having the bank dataset and based on your knowledge in business modeling and analytics, you are required to:

  • Using XLMiner, build a model that can predict if a customer will subscribe for a term deposit or not.
  • Provide a report to the bank management, briefly outlining the process of model development in the context of the CRISP-DM methodology. You are to

justify your model proposal and give arguments that would convince the bank management that it is the model they are looking for. You can also briefly explain how the selected modeling technique works in order to provide the prediction required. Should you have any additional findings or insights that you believe the bank management should know, you can include them in your report.

WhatsApp icon