Table 1 Accuracy on real-world datasets. Standard deviations are provided from runs with seeds from 1 to 10
From: Modeling teams performance using deep representational learning on graphs
Dataset | IMDb | Dribbble | Kaggle | |
|---|---|---|---|---|
Datasets | #teams | 586 | 769 | 1013 |
#classes | 3 | 3 | 3 | |
#nodes | 4802 | 5196 | 4183 | |
#edges | 25632 | 304315 | 17789 | |
Class distr. | 33/33/33 | 33/45/22 | 33/33/33 | |
Classical ML | LR | 64.0 ± 5.0 | 63.2 ± 2.5 | 47.3 ± 2.5 |
SVM | 63.3 ± 5.2 | 63.6 ± 2.8 | 46.7 ± 1.5 | |
RF | 63.2 ± 3.8 | 64.4 ± 3.9 | 47.0 ± 2.1 | |
XGBoost | 64.2 ± 4.0 | 64.6 ± 3.3 | 46.5 ± 2.1 | |
MLP | 64.2 ± 3.8 | 65.6 ± 4.5 | 47.2 ± 1.5 | |
GNN | SubGNN | 63.3 ± 3.1 | 62.1 ± 4.1 | − |
MENTOR—T | 66.9 ± 7.1 | 62.1 ± 4.1 | 45.1 ± 1.8 | |
MENTOR—C | 51.7 ± 4.5 | 63.0 ± 3.8 | 46.6 ± 2.1 | |
MENTOR—L | 50.0 ± 3.7 | 45.2 ± 2.4 | 45.4 ± 2.7 | |
MENTOR | 69.1 ± 4.6 | 66.1 ± 3.2 | 45.3 ± 3.3 |