Reducing the rate of loan defaults is of a great importance for lending institutions. A bank wishes to benefit from the data that it must identify major factors affecting the repayment capacity. A dataset including information on 1500 customers was available. The considered variables are reported in the table below. Age in years Level of education Years with current employer Years at current address Household income in thousands Debt to income ratio (x100) Credit card debt in thousands Other debt in thousands Default a. Assume you are a senior manager of the Credit department. What would be your business objectives? How using data analytics can support these objectives? Elaborate b. The analytics team in the bank adopted the logistic regression approach to develop a model identifying the profile of people who are more likely to default. The available dataset was partitioned into training (70%) and validation data. A logistic model was 3 estimated, and the results are reported in the Appendix 4. What predictors are significant? Justify c. Interpret the coefficients and odds ratios of the significant predictors d. Discuss the prediction performance of this model. Would you recommend implementing this model? Explain e. What would be your strategic recommendations to the bank based on your data analysis?

Appendix 4 Variables in the Equation 95% C.I.for EXP(B) B S.E. Wald df Sig ExD(B) Lower Upper Step 1" Age in years -.019 018 837 .333 982 245 1.019 Level of education 003 076 00 1 989 1.003 863 1.165 Years with current -.233 028 70.536 .DO0 .792 .750 837 employer Years at current -.055 .043 1.626 202 .947 .870 1.030 address Household income in -.006 005 1.834 .178 .904 .985 1.003 thousands Debt to income ratio 076 .020 14.481 1 1.078 1.037 1.121 (x100) Credit card debt in 533 074 51.341 1 .000 1.704 1.473 1.971 thousands Other debt in 081 .030 2.404 .114 1.063 085 1.148 thousands Constant -. 145 480 1 753 865 Classification Table (Confusion Matrix) Predicted Selected Cases Unselected Cases Previously defaulted Percentage Previously defaulted Percentage Observed No Yes Correct No Yes Correct Previously No 579 101 85.1 215 57 70.0 defaulted Yes 156 246 61.2 48 08 67.1 Overall Percentage 78.2 74.9