Table 1 Feature selection result for synthetic datasets
From: Interpretable instance disease prediction based on causal feature selection and effect analysis
Dataset | Dataset1 | Dataset2 | Dataset3 | |||
|---|---|---|---|---|---|---|
Metrics (%) | TPR | FDR | TPR | FDR | TPR | FDR |
Our model | 100 | 0 | 100 | 0 | 92 | 0 |
LIME | 13.8 | 86.2 | 100 | 0 | 98.1 | 1.9 |
Shapley | 60.4 | 39.6 | 93.3 | 6.7 | 65.2 | 9.1 |