Table 2 Results for the Random Forest algorithm using different training set sizes and using the same test set (20%) for all the cases, based on their F-Measure, AUC-ROC, and AU-PRC
From: Semi-supervised incremental learning with few examples for discovering medical association rules
Supervised module | |||
|---|---|---|---|
Train/test % | F-measure | AUC-ROC | AU-PRC |
5/20 | 0.63 | 0.67 | 0.68 |
10/20 | 0.64 | 0.67 | 0.68 |
20/20 | 0.66 | 0.68 | 0.69 |
30/20 | 0.66 | 0.69 | 0.71 |
40/20 | 0.67 | 0.70 | 0.72 |
50/20 | 0.66 | 0.71 | 0.70 |
60/20 | 0.68 | 0.71 | 0.73 |
70/20 | 0.70 | 0.72 | 0.74 |
80/20 | 0.71 | 0.73 | 0.74 |