Abstract:With the recent advances in deep learning, data-driven methods have shown compelling performance in various application domains enabling the Smart Cities paradigm. Leveraging spatial–temporal data from multiple sources for (citywide) traffic forecasting is a key to strengthen the smart city management in areas such as urban traffic control, abnormal event detection, etc. Existing approaches of traffic flow prediction mainly rely on the development of various deep neural networks –e.g., Convolutional Neural Networks such as ResNet are used for modeling spatial dependencies among different regions, whereas recurrent neural networks are increasingly implemented for temporal dynamics modeling. Despite their advantages, the existing approaches suffer from limitations of intensive computations, lack of capabilities to properly deal with missing values, and simplistic integration of heterogeneous data. In this paper, we propose a novel urban flow prediction framework by generalizing the hidden states of the model with continuous-time dynamics of the latent states using neural ordinary differential equations (ODE). Specifically, we introduce a discretize-then-optimize approach to improve and balance the prediction accuracy and computational efficiency. It not only guarantees the prediction error but also provides high flexibility for decision-makers. Furthermore, we investigate the factors, both intrinsic and extrinsic, that affect the city traffic volume and use separate neural networks to extract and disentangle the influencing factors, which avoids the brute-force data fusion in previous works. Extensive experiments conducted on the real-world large-scale datasets demonstrate that our method outperforms the state-of-the-art baselines, while requiring significantly less memory cost and fewer model parameters.

Inferring high-resolutional urban flow with internet of mobile things

UrbanFM: Inferring Fine-Grained Urban Flows

Urban Flow Prediction with Spatial–temporal Neural ODEs

Enhancing Urban Flow Maps Via Neural ODEs.

Fine-grained Urban Flow Inference with Unobservable Data Via Space-Time Attraction Learning

Network-wide Traffic Flow Estimation with Insufficient Volume Detection and Crowdsourcing Data

Mobility Viewer: an Eulerian Approach for Studying Urban Crowd Flow

DiffUFlow: Robust Fine-grained Urban Flow Inference with Denoising Diffusion Model

High-Resolution Urban Flows Forecasting with Coarse-Grained Spatiotemporal Data

Urban Crowd Flow Forecasting Based on Cellular Network

Learning to Generate Temporal Origin-destination Flow Based on Urban Regional Features and Traffic Information

Traffic Flow Prediction Based on Deep Learning in Internet of Vehicles

Fine-Grained Urban Flow Inference Via Normalizing Flow (student Abstract).

Predicting Citywide Crowd Flows Using Deep Spatio-Temporal Residual Networks

Dynamic Origin‐destination Flow Estimation Using Automatic Vehicle Identification Data: A 3D Convolutional Neural Network Approach

AdaTM: Fine-grained Urban Flow Inference with Adaptive Knowledge Transfer Across Multiple Cities

Road Network Guided Fine-Grained Urban Traffic Flow Inference

Diffusion Probabilistic Modeling for Fine-Grained Urban Traffic Flow Inference with Relaxed Structural Constraint

Enhancing Fine-Grained Urban Flow Inference Via Incremental Neural Operator

Learn, Assign, and Search: Real-Time Estimation of Dynamic Origin-Destination Flows Using Machine Learning Algorithms

Improving Urban Crowd Flow Prediction on Flexible Region Partition