Abstract:Passenger flow prediction has drawn increasing attention in the deep learning research field due to its great importance in traffic management and public safety. The major challenge of this essential task lies in multiple spatiotemporal correlations that exhibit complex non-linear correlations. Although both the spatial and temporal perspectives have been considered in modeling, most existing works have ignored complex temporal correlations or underlying spatial similarity. In this paper, we identify the unique spatiotemporal correlation of urban metro flow, and propose an attention-based deep spatiotemporal network with multi-task learning (ADST-Net) at a citywide level to predict the future flow from historical observations. ADST-Net uses three independent channels with the same structure to model the recent, daily-periodic and weekly-periodic complicated spatiotemporal correlations, respectively. Specifically, each channel uses the framework of residual networks, the rectified block and the multi-scale convolutions to mine spatiotemporal correlations. The residual networks can effectively overcome the gradient vanishing problem. The rectified block adopts an attentional mechanism to automatically reweigh measurements at different time intervals, and the multi-scale convolutions are used to extract explicit spatial relationships. ADST-Net also introduces an external embedding mechanism to extract the influence of external factors on flow prediction, such as weather conditions. Furthermore, we enforce multi-task learning to utilize transition passenger flow volume prediction as an auxiliary task during the training process for generalization. Through this model, we can not only capture the steady trend, but also the sudden changes of passenger flow. Extensive experimental results on two real-world traffic flow datasets demonstrate the obvious improvement and superior performance of our proposed algorithm compared with state-of-the-art baselines.

Transformer Based Spatial-Temporal Fusion Network for Metro Passenger Flow Forecasting.

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

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

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

Dynamic Relational Graph Convolutional Network for Metro Passenger Flow Forecasting

Spatial-Temporal Attention Fusion Network for short-term passenger flow prediction on holidays in urban rail transit systems

Spatial-Temporal Fusion Graph Neural Networks for Traffic Flow Forecasting

Learning spatial-temporal dynamics and interactivity for short-term passenger flow prediction in urban rail transit

Multi-STGCnet: A Graph Convolution Based Spatial-Temporal Framework for Subway Passenger Flow Forecasting

Multi-level spatial-temporal fusion neural network for traffic flow prediction

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

A multi-feature spatial–temporal fusion network for traffic flow prediction

Spatial–Temporal Fusion Gated Transformer Network (STFGTN) for Traffic Flow Prediction

Multi-Stage Fusion Framework for Short-Term Passenger Flow Forecasting in Urban Rail Transit Systems Using Multi-Source Data

SPATIOTEMPORAL GRAPH CONVOLUTIONAL NEURAL NETWORKS FOR METRO FLOW PREDICTION

ADST: Forecasting Metro Flow Using Attention-Based Deep Spatial-Temporal Networks with Multi-Task Learning

Spatiotemporal Fusion Transformer for large-scale traffic forecasting

Deep Learning for Metro Short-Term Origin-Destination Passenger Flow Forecasting Considering Section Capacity Utilization Ratio

Metro passenger flow forecasting though multi-source time-series fusion: An ensemble deep learning approach

Multi-sequence spatio-temporal feature fusion network for peak-hour passenger flow prediction in urban rail transit

Spatial-temporal multi-factor fusion graph neural network for traffic prediction