Table 7 Computational efficiency comparisons on the Hangzhou metro dataset.
From: TSTA-GCN: trend spatio-temporal traffic flow prediction using adaptive graph convolution network
Model | PVCGN | MGT | SARGCN | TSTA-GCN |
|---|---|---|---|---|
Parameter amount | \(37.6\times 10^6\) | \(2.82\times 10^6\) | \(1.32\times 10^6\) | \(5.49\times 10^6\) |
Average training time | 22.8s/epoch | 30.1s/epoch | 11.2s/epoch | 59.4s/epoch |
