STG2Seq: Spatial-temporal Graph to Sequence Model for Multi-step Passenger Demand Forecasting
This work addresses the problem of accurate passenger demand prediction for vehicle sharing services, which is incremental as it builds on existing graph-based methods with a hierarchical structure.
The authors tackled multi-step passenger demand forecasting for on-demand vehicle sharing services by proposing STG2Seq, a spatial-temporal graph to sequence model that captures nonlinear and dynamic dependencies, and demonstrated its superior performance over baseline and state-of-the-art models on three real-world datasets.
Multi-step passenger demand forecasting is a crucial task in on-demand vehicle sharing services. However, predicting passenger demand over multiple time horizons is generally challenging due to the nonlinear and dynamic spatial-temporal dependencies. In this work, we propose to model multi-step citywide passenger demand prediction based on a graph and use a hierarchical graph convolutional structure to capture both spatial and temporal correlations simultaneously. Our model consists of three parts: 1) a long-term encoder to encode historical passenger demands; 2) a short-term encoder to derive the next-step prediction for generating multi-step prediction; 3) an attention-based output module to model the dynamic temporal and channel-wise information. Experiments on three real-world datasets show that our model consistently outperforms many baseline methods and state-of-the-art models.