Table 3 Comparison of base classifiers; LR, SVM (with RBF kernel), RF, and their ensembles from the extracted feature from GCN last layer using METABRIC dataset.
From: Breast cancer survival prognosis using the graph convolutional network with Choquet fuzzy integral
Metrics | Acc | Mcc | Pre | Sn | Sp | Bal_Acc | F1-Measure |
|---|---|---|---|---|---|---|---|
GCN-RF | 0.797 | 0.436 | 0.609 | 0.530 | 0.885 | 0.707 | 0.563 |
GCN-RBF | 0.778 | 0.490 | 0.539 | 0.751 | 0.774 | 0.761 | 0.627 |
GCN-LR | 0.783 | 0.492 | 0.547 | 0.741 | 0.777 | 0.759 | 0.628 |
GCN-Majority Voting | 0.786 | 0.490 | 0.552 | 0.725 | 0.806 | 0.765 | 0.626 |
ChoqFuzGCN | 0.820 | 0.528 | 0.630 | 0.666 | 0.871 | 0.769 | 0.647 |
