Tutorial_02_tasks (1)
QBUS2820 – Predictive Analytics
¶Tutorial 2 Tasks¶
Task 1 – Regression Modelling¶
How do customer attributes and marketing efforts affect spending?¶
In order to optimally target customers the sales and marketing teams want to estimate how much a particular customer will spend, given a set of attributes.
Using “DirectMarketing.xlsx” data, inspect and visualize the data to investigate which variables do you think would be helpful to explain output variable ‘AmountSpent’?
Build a linear regression model with dependant variable AmountSpent and independant variables Salary, Children, Gender and Catalogs.
Which variables are significant and which are not?
What is the $R^2$ value of the model? How could you potentially improve the model?
To test your model, you would like to predict sales amount with following features: 50,000 salary, 0 children, female and 12 sent catalogs.
Predict on multiple people at the same time (make up a few people that you might be interested in, does the result make sense to you?)