Abstract:Accurate prediction of citywide short-term metro passenger flow is essential to urban management and transport scheduling. Recently, an increasing number of researchers have applied deep learning models to passenger flow prediction. Nevertheless, the task is still challenging due to the complex spatial dependency on the metro network and the time-varying traffic patterns. Therefore, we propose a novel deep learning architecture combining graph attention networks (GAT) with long short-term memory (LSTM) networks, which is called the hybrid GLM (hybrid GAT and LSTM Model). The proposed model captures the spatial dependency via the graph attention layers and learns the temporal dependency via the LSTM layers. Moreover, some external factors are embedded. We tested the hybrid GLM by predicting the metro passenger flow in Shanghai, China. The results are compared with the forecasts from some typical data-driven models. The hybrid GLM gets the smallest root-mean-square error (RMSE) and mean absolute percentage error (MAPE) in different time intervals (TIs), which exhibits the superiority of the proposed model. In particular, in the TI 10 min, the hybrid GLM brings about 6–30% extra improvements in terms of RMSE. We additionally explore the sensitivity of the model to its parameters, which will aid the application of this model.

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve This paper aims to address the issue of short-term prediction of subway passenger flow within urban areas. Accurate prediction of subway passenger flow is crucial for urban management and traffic scheduling. However, this task remains challenging due to the complex spatiotemporal dependencies within the subway network and the varying traffic patterns over time. Specifically, existing methods have the following shortcomings: 1. **Statistical Methods**: Although simple, they fail to capture nonlinear features. 2. **Machine Learning Methods**: While they improve upon the shortcomings of statistical methods, they still cannot handle raw spatiotemporal data and require complex manual feature engineering. 3. **Deep Learning Methods**: Although capable of automatically extracting features and performing end-to-end predictions, they still have limitations when dealing with complex topological structures in non-Euclidean spaces. Therefore, this paper proposes a hybrid deep learning model (Hybrid GLM) that combines Graph Attention Networks (GAT) and Long Short-Term Memory Networks (LSTM) to overcome the aforementioned issues. The model captures spatial dependencies through the graph attention layer and learns temporal dependencies through the LSTM layer. Additionally, it embeds external factors such as weather conditions, air quality, and holidays to improve prediction accuracy. ### Main Contributions 1. **Proposed a Hybrid Graph Structure Model**: This model can capture complex topological dependencies, particularly addressing the issue that traditional Graph Convolutional Networks (GCN) cannot be applied to asymmetric networks. 2. **Constructed a New Framework**: This framework jointly models the spatial, dynamic temporal, and external dependencies of subway passenger flow data. 3. **Experimental Validation**: Extensive experiments were conducted using real-world datasets, and the results show that the model reduces prediction errors by approximately 6% to 10% compared to the best baseline models at different time intervals. 4. **Practical Application**: The model helps urban managers optimize traffic efficiency, improve the quality of subway services, and provides new opportunities for the construction of intelligent transportation systems. Through these contributions, this paper provides an effective and accurate solution for the short-term prediction of urban subway passenger flow.

A Hybrid GLM Model for Predicting Citywide Spatio-Temporal Metro Passenger Flow

Hybrid Approach Combining Modified Gravity Model and Deep Learning for Short-Term Forecasting of Metro Transit Passenger Flows

A dynamic multigraph and multidimensional attention neural network model for metro passenger flow prediction

MGSTCN: A Multi-Graph Spatio-Temporal Convolutional Network for Metro Passenger Flow Prediction

A Hybrid Deep Learning Approach for Real-Time Estimation of Passenger Traffic Flow in Urban Railway Systems

A dynamic graph deep learning model with multivariate empirical mode decomposition for network‐wide metro passenger flow prediction

Fuzzy Hybrid Framework with Dynamic Weights for Short‐term Traffic Flow Prediction by Mining Spatio‐temporal Correlations

A method for short-term passenger flow prediction in urban rail transit based on deep learning

Optimizing Metro Passenger Flow Prediction: Integrating Machine Learning and Time-Series Analysis with Multimodal Data Fusion

Multi-Spatio-Temporal Convolutional Neural Network for Short-Term Metro Passenger Flow Prediction

A Hybrid Deep Learning Model for Short-Term Traffic Flow Pre-Diction Considering Spatiotemporal Features

Short-Term Metro Origin-Destination Passenger Flow Prediction via Spatio-Temporal Dynamic Attentive Multi-Hypergraph Network

Short-Term Inbound and Outbound Passenger Flow Prediction for New Metro Stations Based on Clustering and Deep Learning

A geographic-semantic context-aware urban commuting flow prediction model using graph neural network

GC-LSTM: A Deep Spatiotemporal Model for Passenger Flow Forecasting of High-Speed Rail Network.

Short-term Passenger Flow Prediction for Hangzhou Metro based on Machine Learning Algorithms

Empirical mode decomposition based long short-term memory neural network forecasting model for the short-term metro passenger flow

IPF-HMGNN: A novel integrative prediction framework for metro passenger flow

Short-Term Online Forecasting for Passenger Origin–Destination (OD) Flows of Urban Rail Transit: A Graph–Temporal Fused Deep Learning Method

Deep-learning Architecture for Short-term Passenger Flow Forecasting in Urban Rail Transit

Learning Spatial-Temporal Dynamics for Short-Term Passenger Flow Prediction in Urban Rail Transit