With reference to the above case, describe the pros and cons of using a metric accuracy vs. recall vs. Area-Under-Curve for your classification modelling.

ASSIGNMENT

QUESTION 1 25 MARKS

With reference to the above case, answer all of the following questions:

a) Based on the business challenges that Auto Finance Ltd. is facing, list and explain a segmentation problem and a classification problem respectively.

b) Select an analytics model for each problem you have identified in Q1a and justify your choice.

c) Outline relevant variables for each analytics model you have identified in Q1(b). Justify your choice of these variables.

QUESTION 2 15 MARKS

Explain how each of your proposed modelling techniques in Question 1 b) can help Auto Finance Ltd. generate insights about a potential customer defaulting on a loan.

QUESTION 3 10 MARKS

With reference to the above case, describe the pros and cons of using a metric accuracy vs. recall vs. Area-Under-Curve for your classification modelling. Justify which metric is the best based on the context of the case.

QUESTION 4 15 MARKS

There are concerns about the ethical issues involved in scoring a customer’s credits.

Identify three ethical issues that Auto Finance Ltd. could face. Provide your recommendations to mitigate each ethical issue identified.

With reference to the above case, describe the pros and cons of using a metric accuracy vs. recall vs. Area-Under-Curve for your classification modelling.