Traffic Speed Prediction Based on LSTM-Graph Attention Network (L-GAT)

摘要

In recent years, with the rise of graph convolutional neural networks, traffic network prediction has become a hot topic. Affected by this, this paper proposes a new traffic network speed prediction method, named LSTM-Graph Attention Network (L-GAT). This method can not only capture the spatial characteristics of the traffic network, but also learn the time dynamics. In addition, we conducted experiments on the public data set of Didi, which proved that the L-GAT method is superior to other methods in traffic speed prediction.