The automobile insurance company would like to learn how the prediction models work. Therefore, the data science team also built an explainable prediction model and obtained the following prediction probabilities on a small test set. The first column provides the Claim ID. The second column provides the prediction probabilities obtained from the model. For instance, Claim 111 is predicted as legitimate with probability 0.85 and not legitimate (or default) with probability 0.15. The third column presents the actual outcome given by the test data, respectively. Claim ID Probability for legitimate and not legitimate prediction Actual Target Value Legitimate Not Legitimate 111 (0.85, 0.15) 112 (0.53, 0.47) 113 (0.75, 0.25) 114 (0.45, 0.55) Legitimate Legitimate 115 (0.10, 0.90) Not Legitimate 116 (0.65, 0.35) Not Legitimate 117 (0.80, 0.20) 118 (0.62, 0.48) Legitimate Legitimate 119 (0.73, 0.37) Legitimate 120 (0.37, 0.63) Not Legitimate