Which metrics will indicate an ML solution that will provide the GREATEST probability of detecting an abnormality?
Precision = 0.91 -Recall = 0.6
Precision = 0.61 -Recall = 0.98
Precision = 0.7 -Recall = 0.9
Precision = 0.98 -Recall = 0.8
Explanations:
Precision is high at 0.91, indicating that when an abnormality is predicted, it is likely to be correct. However, the recall is low at 0.6, meaning that many actual abnormalities are missed. This results in a low detection rate.
Precision is low at 0.61, indicating that a significant number of false positives occur. However, the recall is very high at 0.98, meaning that nearly all actual abnormalities are detected. This maximizes the likelihood of identifying issues.
While both precision (0.7) and recall (0.9) are reasonably good, they do not reach the levels seen in option B. Although it detects most abnormalities, the lower precision means that a higher number of false positives could occur.
Precision is very high at 0.98, suggesting excellent accuracy in predictions of abnormalities. However, the recall at 0.8 indicates that not all actual abnormalities are detected, potentially leading to missed issues despite high precision.