Spatiotemporal Virtual Graph Convolution Network for Key Origin-Destination Flow Prediction in Metro System

<div>Parameter tuning results. (a) Variation of predicted performance under different epochs. (b) Variation of predicted performance under different hidden units.</div>

Mathematical Problems in Engineering

fig4

Figure 4

Figure 4: Spatiotemporal Virtual Graph Convolution Network for Key Origin-Destination Flow Prediction in Metro System 