Question 5: Decision Tree (2+2+2+2+2 marks) The output below comes from running the rpart function in the R Statistical Environment on the car insurance dataset just described. 1) root 5000 1250 0 (0.75 0.25) 2) female>=0.5 2530 506 0 (0.80, 0.20) 4) lowSES < 0.5 1870 1500 (0.92 0.08) 5) lowSES>=0.5 660 231 0 (0.65 0.35) 3) female < 0.5 2470 741 0 (0.70 0.30) 6) age> 31.5 1820 455 0 (0.75 0.25) 7) age < 31.5 650 299 0 (0.54 0.46) * * * 14) turbo < 0.5 470 141 0 (0.70 0.30) 15) turbo>=0.5 180 45 1 (0.25 0.75) * * (A) List the terminal nodes (also called leaves) from the Rpart classification tree. (B) Calculate the entropy improvement from the first split. (C) Produce the confusion matrix for the classification tree. (D) Calculate the averaged class error for the classification tree. (E) Which node does the customer with ID 2 belong to? For a new customer with the same values of the female, age, turbo and low_SES attributes, do you predict that they will make a large claim within 12 months?