Short-term Subway Passenger Flow Prediction based on GCN-LSTM-MHA

The short-term prediction of subway passenger flow is of great reference value for the operation and management of the subway system. To improve the accuracy of the prediction results, this paper designs a combined model based on Graph Convolutional Neural Network and Long Short-Term Memory Network with Multi-Head Attention mechanism. Moreover, the Tree-Structured Parzen Estimator is used to optimize the relatively important hyperparameters in the model. Experimental results demonstrate that the GCN-LSTM-MHA model effectively captures the spatiotemporal correlation of subway passenger flow data and outperforms other baseline models in terms of prediction accuracy.