The following table shows ten testing data points and the score assigned to each by a classifier. (In logistic regression, the score is the probability output of data being positive; In decision tree, the score is the positive data proportion in the corresponding leaf node). Draw the ROC curve for this classifier and write down detailed steps. For each decision threshold, you need to write down the confusion matrix as well as the corresponding true positive rate and false positive rate. You have to do it manually instead of using Python. Hint: You should calculate the FPR and TPR under all possible different decision thresholds, plot and connect them together on the graph. By threshold x, we are saying if the probability estimate >= x, we predict it as p; otherwise, we predict it as n. Actual Class Inst# Score 1 p 0.9 2 p 0.78 3 n 0.66 4 p 0.54 5 n 0.506 6 p 0.4 7 p 0.38 8 n 0.3 9 p 0.15 10 n 0.1