Data mining

Q26

Question 26 6 Points Consider the following confusion matrix of a system powered by a DM used by a bank to predict whether a loan applicant is going to pay back the loan or not. The table shows the performance of the model through a confusion matrix for 100 cases. Predicted Accept Reject 30 10 Actual Accept Reject O 60 Calculate the following measures: The accuracy of this model: Blank 1 %. Recall of the model: Blank 2 %. The specificity of the model: Blank 3 %. Precision of the model: Blank 4 %. Prevalence* of the model: Blank 5 %. The Lift of this mode: Blank 6. Notes: *Prevalence refers to how often does the (+ve) condition actually occur in the dataset. For the above measures strictly write the results in an integer format (e.g.:76) without a decimal point, except the Lift, write it in this format (X.X), (e.g. 3.6) Don't use ordinary fractions (e.g if the answer is 50%, don't write it like: 1/2. rather write it as 50. Don't put the ( % ) sign in the blank, I repeat DON'T PUT the ( % ) sign in the blank, it is already there outside the blank