Table 22 Precision of explanations of AdaBoost SAMME.R
From: Ada-WHIPS: explaining AdaBoost classification with applications in the health sciences
Data | Ada-WHIPS | Anchors | LORE |
|---|---|---|---|
Breast cancer | 0.9831±0.0014 | 0.9793±0.0103 | 0.8215±0.0210 |
Cardiotocography | 0.9324±0.0032 | 0.9117±0.0107 | 0.7931±0.0110 |
Diabetic retinopathy | 0.8272±0.0073 | 0.8164±0.0175 | 0.5481±0.0203 |
Cleveland heart | 0.9059±0.0105 | 0.9640±0.0189 | 0.5971±0.0293 |
Mental health survey ’16 | 0.9849±0.0013 | 0.9502±0.0100 | 0.9129±0.0124 |
Mental health survey ’14 | 0.9030±0.0043 | 0.9811±0.0056 | N/A |
Hospital readmission | 0.9129±0.0013 | 0.9811±0.0032 | N/A |
Thyroid | 0.9481±0.0015 | 0.8154±0.0110 | N/A |
Understanding society | 0.8677±0.0043 | 0.8903±0.0081 | N/A |