Table 1 below summarizes the prediction scores of a model on a dataset with 5 observations. Please fill out Table 2 and draw the ROC curse of this model. It is fine to draw it on paper and then take a picture of it. After that, fill out Table 3 and check if the percentage of concordant (ranking-consistent) positive-negative pairs is the same as AUC of the ROC curve.

Table 1. Prediction Scores and Actual outcomes

Instance	Actual Outcome	Prediction Score
A	1	0.8
B	0	0.6
C	1	0.6
D	1	0.5
E	0	0.2

Table 2. True Positive Rate and False Positive Rate at Different Thresholds

Threshold	Predicted Positive Instances	% True Positive	% False Positive
(0.8, )	{}	0%	0%
(0.6, 0.8]	{A}	50%	0%
(0.5, 0.6]
(0.2, 0.5]
(,0.2]

Note: squared bracket is inclusive, whereas parenthesis is exclusive.

Table 3. Percentage of Concordant Positive-Negative Pairs

Positive-Negative Pair	Ranking Consistent?	% Consistent

Note: You may need to add more rows into this table.