Table 2 Model performances predicting patient avoidance of refried beans and beer. (a) Accuracy and AUC for each model against the test holdout set. (b) Confusion matrices for each model on the test holdout set.
From: The risk of shortcutting in deep learning algorithms for medical imaging research
a) | ||||
---|---|---|---|---|
Refried beans | Beer | |||
Accuracy | 0.600 | 0.702 | ||
AUC | 0.631 | 0.734 |
b) | ||||
---|---|---|---|---|
True answer | Predicted answer | |||
Refried beans | Beer | |||
Yes | No | Yes | No | |
Yes | 1,258 | 686 | 1,876 | 610 |
No | 839 | 1,031 | 529 | 810 |