Table 5 Macro-F1 scores for classification of nodes in epidemic states according to different SIR epidemic processes for synthetic datasets. For each \((\beta ,\mu )\) we highlight the best score
From: Time-varying graph representation learning via higher-order skip-gram with negative sampling
(β,μ) | Model | Dataset | |
|---|---|---|---|
OpenABM-2k-100 | OpenABM-5k-20 | ||
(0.25,0.002) | DyANE | 57.9 ± 1.8 | 59.6 ± 1.7 |
\(\text{HOSGNS} ^{(\text{stat})}\) | 31.2 ± 0.1 | 27.8 ± 0.6 | |
\(\text{HOSGNS} ^{(\text{dyn})}\) | 57.5 ± 1.8 | 61.0 ± 1.1 | |
(0.0625,0.002) | DyANE | 61.8 ± 0.4 | 53.8 ± 1.3 |
\(\text{HOSGNS} ^{(\text{stat})}\) | 29.8 ± 0.2 | 29.4 ± 1.4 | |
\(\text{HOSGNS} ^{(\text{dyn})}\) | 59.5 ± 0.9 | 54.5 ± 1.4 | |
(0.1875,0.001) | DyANE | 60.3 ± 1.4 | 59.6 ± 1.5 |
\(\text{HOSGNS} ^{(\text{stat})}\) | 31.9 ± 0.2 | 27.4 ± 0.7 | |
\(\text{HOSGNS} ^{(\text{dyn})}\) | 60.5 ± 1.1 | 60.9 ± 1.0 | |