Question abcd

e 1: Suppose that we are looking at predicting whether credit card transactions are valid or fraudulent. Based upon past history we developed a classification model and applied it to 10,000 known transactions with the following result. Valid Fraudulent Total 2: Approved 9800 20 9820 Rejected 100 80 180 Total 9900 100 10,000 Suppose that we have the following consequences to each combination. -A valid transaction that is approved generates an average profit of $10 in fee and interest. -A fraudulent transaction that is approved results in an average loss of $300. A fraudulent transaction that is rejected results in $0 profit/loss. -A valid transaction that is rejected results in a loss of goodwill of $20. Answer the following questions in order: a) (2pts)What is the Accuracy of the model? b) (2pts)What is the Sensitivity of the model? c) (2pts)What is the Specificity of the model? d) (4pts)What is the expected profit per transaction if this model is used? MacBook