Abstract. Forecasting urban metro flow accurately plays an important role for station management and passenger safety. Owing to the limitations of non-linearity and complexity of traffic flow data, traditional methods cannot satisfy the requirements of effectively capturing spatiotemporal dependencies at the metro network level, which makes it difficult to demonstrate high performance. In this paper, a novel deep learning method is proposed based on Graph Neural Networks (GNN), named STGCN-Metro (SpatioTemporal Graph Convolutional Network based on Metro network), to forecast the short-term inflow and outflow volumes of metro passengers. The proposed model is composed of two spatiotemporal convolutional blocks, which is integrated with the Dilated Convolutional Neural Network (DCNN) and Cluster-Graph Convolutional Network (Cluster-GCN). The DCNN is employed with different dilation rates to capture temporal dependence in larger receptive field. In addition, compare with GCN, the Cluster-GCN is applied the graph clustering algorithms to reduce computational resources considering spatial heterogeneity. A real-world dataset collected in Shanghai metro stations is conducted for validation, and the results demonstrate that the proposed model achieves higher performance, outperforming some well-known baseline models.

Abstract. Accurate crime prediction plays an important role in public safety, providing technical guidance and decision support for the police and government departments. Due to the dynamics and imbalance of crime distribution, it is difficult to build predictive models for it. Specifically, the fine-grained and non-linear spatiotemporal dependencies of crime data cannot be captured accurately. In this paper, a neural network model ST-ACLCrime based on ConvLSTM and SE block was proposed to predict the number of theft crimes in hotspot areas. By overlaying ConvLSTM layers, fine-grained spatiotemporal dependencies are captured while preserving spatial location information. To further enhance the global channel feature representation, SE block is used to recalibrate the channel features and enhance the channel inter-dependencies. In addition, the closeness and the period components are set to dynamically capture the dependence of different time trends. We choose the city of Chicago as the study case, and use a multi-level spatial grid to divide the whole city area. The experimental results show that the proposed model exceeds all baseline model, such as HA, CNN, LSTM, CNN-LSTM and ConvLSTM. It was effectively capturing spatiotemporal dependence and improving prediction accuracy.

No abstract