Table 1 Accuracy (in %) of the different nested cross-validation (NCV) and generalization tests depending on the training dataset
From: Machine learning on multiple topological materials datasets
Test\Train | M | T | M\T | M ∩ T | T\M | (M\T) ∪ (T\M) | M ∪ T |
|---|---|---|---|---|---|---|---|
NCV | 85.7 | 80.7 | 84.1 | 85.1 | 72.1 | 79.9 | 82.9 |
M\T | 84.8* | 80.3 | 84.1* | 78.9 | 77.1 | 85.6* | 85.8* |
M ∩ T | 86.1* | 86.0* | 82.0 | 85.1* | 81.6 | 83.6 | 86.6* |
T\M | 71.8 | 73.2* | 69.8 | 70.2 | 72.1* | 73.3* | 74.3* |
M ∪ T | 81.8 | 80.6 | 79.3 | 79.0 | 77.5 | 81.4 | 82.9 |