2 Classification Metrics You have trained a Logistic Regression classifier which gives you the following predictions on a test set: (a) (5 points) Draw the Receiver Operating Characteristic (ROC) curve, Show clearly the points where the curve changes direction by e.g. including ticks on the x - and y-axis in the corresponding locations. (b) (2 points) Compute the AUC score. (c) (2 points) Draw the confusion matrix when the decision threshold is 0.5. (d) (2 points) Using the previous confusion matrix, compute Accuracy, Precision, Recall, and F1 score, (e) (5 points) Can we improve any of the previous scores (without a negative effect on any of the other scores) by changing the threshold? If yes, which threshold value would you choose and why? If not, explain why not