Awesome Traffic Prediction

Chinese version: README.zh.md

This repository contains useful resources for traffic prediction, including popular papers, datasets, tutorials, toolkits, and other helpful repositories.

Visual Highlights

Latest monthly title-term word cloud:

Latest Top-K term frequency chart:

Yearly hotspot comparison:

Yearly theme trend lines:

• 0x00 Papers
  • Reviews
  • Deep Learning Based Traffic Prediction Methods
  • Curated Digest (Recent Years)
  • Statistic Based Traffic Prediction Methods
  • Time Series Forecasting
  • Temporal Network Embedding
• 0x01 Tutorial
  • Textbook
  • Blogs
• 0x02 DataSource
  • Datasets
  • Trajectory
  • ETA
  • Website
• 0x03 Toolkits
• 0x04 Conferences & Journals
• 0x05 Research Group
  • Company
  • College
• 0x06 Related Repositories
• 0x07 Benchmark & Evaluation
  • Benchmark Matrix
  • Auto Trend Visuals
• Contribution

0x00 Papers

Reviews

1. [TITS 2015] Traffic Flow Prediction With Big Data: A Deep Learning Approach [paper]
2. [KDD 2020] Deep Learning for Spatio-Temporal Data Mining: A Survey [paper]
3. [Information Fusion 2020] Urban flow prediction from spatiotemporal data using machine learning: A survey [paper]
4. [Arxiv 2020] Deep Learning on Traffic Prediction: Methods, Analysis and Future Directions [paper]
5. [Arxiv 2021] Graph Neural Network for Traffic Forecasting: A Survey [paper]
6. [Applied Intelligence 2022] Spatial-temporal graph neural network for traffic forecasting: An overview and open research issues [paper]

Deep Learning Based Traffic Prediction Methods

2015

1. [NIPS 2015] Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting [paper]

2016

1. [Sigspatial 2016] DNN-Based Prediction Model for Spatio-Temporal Data [paper] [code]

2017

1. [AAAI 2017] Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction [paper]
2. [ISPRS 2017] Road2Vec: Measuring Traffic Interactions in Urban Road System from Massive Travel Routes [paper]
3. [Arxiv 2017] DeepTransport: Learning Spatial-Temporal Dependency for Traffic Condition Forecasting [paper]

2018

1. [TITS 2019] T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction [paper] [code]
2. [TITS 2018] Missing Value Imputation for Traffic-Related Time Series Data Based on a Multi-View Learning Method [paper]
3. [IJCAI 2018] Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting [paper] [code] [review]
4. [IJCAI 2018] LC-RNN: A Deep Learning Model for Traffic Speed Prediction [paper]
5. [IJCAI 2018] GeoMAN: Multi-level Attention Networks for Geo-sensory Time Series Prediction [paper]
6. [ICLR 2018] Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting [paper] [code-official-tf] [code-pytorch] [review] [data]
7. [KDD 2018] Hetero-ConvLSTM: A Deep Learning Approach to Traffic Accident Prediction on Heterogeneous Spatio-Temporal Data [paper]
8. [CS224W 2018] Efficient Traffic Forecasting With Graph Embedding [paper] [code]
9. [CVPR 2018] Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks [paper] [paper-ieee] [code]
10. [AAAI 2018] Deep Multi-View Spatial-Temporal Network for Taxi Demand Prediction [paper] [code] [data]
11. ...

2019

1. [TITS 2019] TrafficGAN: Network-Scale Deep Traffic Prediction With Generative Adversarial Nets [paper]
2. [TITS 2019] Contextualized Spatial–Temporal Network for Taxi Origin-Destination Demand Prediction [paper] [code]
3. [TITS 2019] Deep Spatial-Temporal 3D Convolutional Neural Networks for Traffic Data Forecasting [paper]
4. [IJCAI 2019] Graph WaveNet for Deep Spatial-Temporal Graph Modeling [paper] [code]
5. [IJCAI 2019] GSTNet: Global Spatial-Temporal Network for Traffic Flow Prediction [paper]
6. [AAAI 2019] Revisiting Spatial-Temporal Similarity: A Deep Learning Framework for Traffic Prediction [paper] [code]
7. [AAAI 2019] DeepSTN+: Context-aware Spatial-Temporal Neural Network for Crowd Flow Prediction in Metropolis [paper] [code]
8. [AAAI 2019] Attention Based Spatial-Temporal Graph Convolutional Networks for Traffic Flow Forecasting [paper] [code-pytorch]
9. [AAAI 2019] Multi-Range Attentive Bicomponent Graph Convolutional NetworkforTraicForecasting [paper]
10. [WWWC 2019] Learning from Multiple Cities: A Meta-Learning Approach for Spatial-Temporal Prediction [paper]
11. [IWPHM 2019] Spatio-Temporal Clustering of Traffic Data with Deep Embedded Clustering [paper]
12. [ICCV 2019] STGAT: Modeling Spatial-Temporal Interactions for Human Trajectory Prediction [paper] [code]
13. [KDD 2019] Urban Traffic Prediction from Spatio-Temporal Data Using Deep Meta Learning [paper][code]
14. [ICIKM 2019] Matrix Factorization for Spatio-Temporal Neural Networks with Applications to Urban Flow Prediction [paper]
15. [TKDE 2019] Flow prediction in spatio-temporal networks based on multitask deep learning [paper]
16. [IJGIS 2019] Traffic speed prediction for intelligent transportation system based on a deep feature fusion model [paper]
17. [Access 2019] Spatial-Temporal Graph Attention Networks: A Deep Learning Approach for Traffic Forecasting [paper]
18. [Arxiv 2019] Forecaster: A graph transformer for forecasting spatial and time dependent data [paper]
19. [Arxiv 2019] Temporal fusion transformers for interpretable multi-horizon time series forecasting [paper]
20. ……

2020

1. [Arxiv 2020] Spatial-Temporal Transformer Networks for Traffic Flow Forecasting [paper] [code-not-official]
2. [Arxiv 2020] Transfer Learning with Graph Neural Networks for Short-Term Highway Traffic Forecasting [paper] [code]
3. [Arxiv 2020] Bayesian Spatio-Temporal Graph Convolutional Network for Traffic Forecasting [paper]
4. [TGIS 2020] Traffic transformer: Capturing the continuity and periodicity of time series for traffic forecasting [paper]
5. [ICTON 2020] Traffic Prediction in Optical Networks Using Graph Convolutional Generative Adversarial Networks [paper]
6. [AAAI 2020] Spatio-Temporal Graph Structure Learning for Traffic Forecasting [paper] [SOTA]
7. [AAAI 2020] Learning Geo-Contextual Embeddings for Commuting Flow Prediction [paper]
8. [AAAI 2020] Spatial-Temporal Synchronous Graph Convolutional Networks: A New Framework for Spatial-Temporal Network Data Forecasting [paper] [code]
9. [Access 2020] STGAT: Spatial-Temporal Graph Attention Networks for Traffic Flow Forecasting [paper]
10. [Sensor 2020] City-Wide Traffic Flow Forecasting Using a Deep Convolutional Neural Network [paper]
11. [Mobile Computing 2020] BuildSenSys: Reusing Building Sensing Data for Traffic Prediction with Cross-domain Learning [paper]
12. [TKDE 2020] Spatio-Temporal Meta Learning for Urban Traffic Prediction [paper]
13. [WC 2020] What is the Human Mobility in a New City: Transfer Mobility Knowledge Across Cities [paper]
14. [TITS 2020] Traffic Graph Convolutional Recurrent Neural Network: A Deep Learning Framework for Network-Scale Traffic Learning and Forecasting [paper] [code]
15. [NeurIPS 2020] Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting [paper] [code]
16. [AAAI 2020] GMAN: A Graph Multi-Attention Network for Traffic Prediction [paper] [code]
17. [KDD 2020] ConSTGAT: Contextual Spatial-Temporal Graph Attention Network for Travel Time Estimation at Baidu Maps [paper]
18. [KDD 2020] Preserving Dynamic Attention for Long-Term Spatial-Temporal Prediction [paper]
19. [TITS 2020] Temporal Multi-Graph Convolutional Network for Traffic Flow Prediction [paper]
20. [TITS 2020] A Spatial-Temporal Attention Approach for Traffic Prediction [paper]
21. [TITS 2020] Traffic Flow Imputation Using Parallel Data and Generative Adversarial Networks [paper]
22. [WWW 2020] Traffic Flow Prediction via Spatial Temporal Graph Neural Network [paper]
23. [IJGIS 2020] Graph attention temporal convolutional network for traffic speed forecasting on road networks [paper]
24. [Arxiv 2020] ST-GRAT: A Novel Spatio-temporal Graph Attention Network for Accurately Forecasting Dynamically Changing Road Speed [paper]
25. [IF 2020] Spatial Temporal Incidence Dynamic Graph Neural Networks for Traffic Flow Forecasting [paper]
26. [ICDM 2020] TSSRGCN: Temporal Spectral Spatial Retrieval Graph Convolutional Network for Traffic Flow Forecasting [paper]
27. ……

2021

1. [Arxiv 2020] Spatial-Temporal Fusion Graph Neural Networks for Traffic Flow Forecasting [paper] [code]
2. [TITS 2021] Spatial‐temporal attention wavenet: A deep learning framework for traffic prediction considering spatial‐temporal dependencies [paper] [code]
3. [Arxiv 2021] Time Series is a Special Sequence: Forecasting with Sample Convolution and Interaction [paper] [code]
4. [KDD 2021] Dynamic Graph Convolutional Recurrent Network for Traffic Prediction: Benchmark and Solution [paper] [code]
5. [KDD 2021] Dynamic and Multi-faceted Spatio-temporal Deep Learning for Traffic Speed Forecasting [paper]
6. [AAAI 2021] TS2Vec: Towards Universal Representation of Time Series [paper]
7. [Arxiv 2021] Spatio-temporal joint graph convolutional networks for traffic forecasting [paper]
8. [PAKDD 2021] SST-GNN: Simplified Spatio-temporal Traffic forecasting model using Graph Neural Network [paper]
9. [IJCNN 2021] Unified Spatio-Temporal Modeling for Traffic Forecasting using Graph Neural Network [paper]
10. ……

2022

1. [TITS 2022] 2F-TP:Learning Flexible Spatiotemporal Dependency for Flexible Traffic Prediction. [paper]
2. [Arxiv 2022] Pre-training Enhanced Spatial-temporal Graph Neural Network for Multivariate Time Series Forecasting [paper]
3. [Arxiv 2022] A Lightweight and Accurate Spatial-Temporal Transformer for Traffic Forecasting [paper]

2023

1. [Arxiv 2023] Adaptive Graph Convolution Networks for Traffic Flow Forecasting [paper] [code]
2. [ICDE 2023] Dynamic Hypergraph Structure Learning for Traffic Flow Forecasting [paper]

2024

1. [Arxiv 2024] UniST: A Prompt-Empowered Universal Model for Urban Spatio-Temporal Prediction [paper]
2. [Arxiv 2024] COOL: A Conjoint Perspective on Spatio-Temporal Graph Neural Network for Traffic Forecasting [paper]
3. [Complex & Intelligent Systems 2024] Generalized spatial–temporal regression graph convolutional transformer for traffic forecasting [paper]
4. [ISPRS IJGI 2024] Traffic Flow Prediction Based on Federated Learning and Spatio-Temporal Graph Neural Networks [paper]

2025

1. [Arxiv 2025] GraphSparseNet: a Novel Method for Large Scale Trafffic Flow Prediction [paper]
2. [Arxiv 2025] Embedding spatial context in urban traffic forecasting with contrastive pre-training [paper] [code]
3. [Sensors 2025] An Adaptive Spatio-Temporal Traffic Flow Prediction Using Self-Attention and Multi-Graph Networks [paper]
4. [Neural Computing and Applications 2025] Short-term multi-step-ahead sector-based traffic flow prediction based on the attention-enhanced graph convolutional LSTM network (AGC-LSTM) [paper]

2026

1. [Neural Networks 2026] Adaptive dynamic spatial-temporal graph convolutional neural network for traffic flow prediction [paper]
2. [Electronics 2026] Dynamic Graph Information Bottleneck for Traffic Prediction [paper]
3. [Scientific Reports 2026] 6G conditioned spatiotemporal graph neural networks for real time traffic flow prediction [paper]
4. [Arxiv 2026] PIMCST: Physics-Informed Multi-Phase Consensus and Spatio-Temporal Few-Shot Learning for Traffic Flow Forecasting [paper]
5. [Expert Systems with Applications 2026] HySRD-Net: Hybrid spectral residual diffusion on spatio-temporal graphs for probabilistic traffic forecasting [paper]
6. [PLoS ONE 2026] Feature-enhanced iTransformer: A two-stage framework for high-accuracy long-horizon traffic flow forecasting [paper]
7. [IET Intelligent Transport Systems 2026] A Novel Attention-Weighted VMD-LSSVM Model for High-Accuracy Short-Term Traffic Prediction [paper]
8. [Electronics 2026] Multi-Scale Graph-Decoupling Spatial-Temporal Network for Traffic Flow Forecasting in Complex Urban Environments [paper]
9. [Systems 2026] PDR-STGCN: An Enhanced STGCN with Multi-Scale Periodic Fusion and a Dynamic Relational Graph for Traffic Forecasting [paper]
10. [Lecture Notes in Computer Science 2026] STVMamba: A Spatio-Temporal-Variable Modeling Network with Selective State Space Mechanism for Long-Horizon Traffic Forecasting [paper]
11. [Electronics 2026] TSAformer: A Traffic Flow Prediction Model Based on Cross-Dimensional Dependency Capture [paper]

Curated Digest (Recent Years)

This digest reorganizes recent traffic flow forecasting papers by task and method tags for faster navigation. Method tags are lightweight manual labels for quick indexing.

Statistic Based Traffic Prediction Methods

2018

1. [TITS 2018] Probabilistic Data Fusion for Short-Term Traffic Prediction With Semiparametric Density Ratio Model [paper]

2019

1. [TRPET 2019] A generalized Bayesian traffic model [paper]

Time Series Forecasting

1. [ICLR 2019] Context-aware Forecasting for Multivariate Stationary Time-series [paper]
2. [Arxiv 2020] Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting [paper] [code]

Temporal Network Embedding

2021

1. [KDD 2021] Discrete-time Temporal Network Embedding via Implicit Hierarchical Learning in Hyperbolic Space

0x01 Tutorial

Textbook

1. Urban Computing
2. Traffic Flow Dynamics: Data, Models and Simulation
3. Transportation Network Analysis
4. Fundamentals of Transportation and Traffic Operations
5. ……

Blogs

1. Traffic prediction with advanced Graph Neural Networks
2. LibCity Documentation (Quick Start)
3. LibCity Tutorials
4. DCRNN Official Repository
5. Traffic-Benchmark (KDD 2021)
6. PyTorch Geometric Temporal Documentation
7. ……

0x02 DataSource

Datasets

Traffic Dataset in Non-China Region

1. Cityscapes

2. New York City

3. NYC Bike

4. NYC Taxi

5. Train Station Dataset

6. Apolloscape

   The trajectory dataset includes camera images, LiDAR point clouds, and manually annotated trajectories collected in Beijing under varying lighting and traffic conditions.

7. data.world.traffic

8. PEMS-SF Dataset From UCI

   Each attribute describes detector occupancy at a station for a given timestamp (0-1). Station IDs are listed in the stations_list file; location metadata (GPS, freeway, direction) is available from PeMS.

9. Seattle Inductive Loop Detector Dataset

10. Road location and traffic data

11. INRIX – Driving Intelligence

    Charged Data

12. Los Angeles (METR-LA)

13. ……

Traffic Dataset in China

1. Baidu Open Data

2. DiDi Gaia Initiative

   An anonymized road condition dataset for Xi'an, including road network topology, road attributes, and historical/real-time traffic flow.

3. HZJTD

   Data collected by the Hangzhou Comprehensive Transportation Research Center, including traffic conditions, speed, and congestion indices on 202 roads in the urban area.

Commute flow

1. Longitudinal Employer-Household Dynamics
2. ……

Point of Interest/Land use

1. PLUTO

   provide landuse data organized in .csv format, including more than 70 fields managed by city insititution.

2. ……

Trajectory

ETA

1. Shenzhen ride-hailing trajectory dataset (SIGSPATIAL 2021 GISCUP ETA)

Website

1. Welcome to PeMS
2. ……

0x03 Toolkits

1. Open Source Routing Machine
2. PyG Temporal
3. LibCity
4. OpenSTL
5. Traffic-Benchmark
6. UrbanPy

0x04 Conferences & Journals

1. ACM SIGSPATIAL SpatialDI
2. IEEE Transactions on Intelligent Transportation Systems
3. Association for the Advancement of Artificial Intelligence

0x05 Research Group

Company

1. http://urban-computing.com/yuzheng
2. DeepMind
3. Waymo Research
4. Baidu Apollo
5. ……

College

1. Beijing Jiaotong University - Dr. Huaiyu Wan's group

   Research Interest: Data mining and information extraction. e.g. Spatio-temporal data mining, social network mining, text information extraction and application of knowledge graph.

2. Tsinghua University - Future Intelligence Lab (FIB)

   Focus: urban science, spatio-temporal modeling, and large-scale AI systems.

3. Tongji University - ITS Research Center (Prof. Xiaoguang Yang)

   Focus: traffic system engineering, traffic control, public transportation systems, and ITS.

4. Tongji University - STEP Lab (Prof. Yuchuan Du)

   Focus: traffic sensing, transport infrastructure intelligence, and AI for transportation systems.

5. Central South University - Institute of Intelligent Measurement, Control and Information Technology

   Focus: sensing, diagnosis, and intelligent collaboration for transportation systems.

6. Central South University - Institute of Intelligent Simulation and Decision for Integrated Transport

   Focus: transport optimization, simulation, and data-driven decision support.

7. Southwest Jiaotong University - Intelligent Design Lab for Complex Transportation Systems

   Focus: data-driven and AI-based methods for complex transportation system design.

8. UC Berkeley - Institute of Transportation Studies (ITS)

   Focus: transportation systems analysis, ITS, traffic operations, and mobility research.

9. UC Berkeley - PATH

   Focus: intelligent transportation systems, connected/autonomous mobility, and corridor management.

10. MIT - Senseable City Lab

    Focus: data-driven urban science, mobility analytics, and real-time city sensing.

11. UCL - Centre for Advanced Spatial Analysis (CASA)

    Focus: urban analytics, spatial-temporal modeling, and computational city science.

12. NTU Singapore - Transport Research Centre (TRC)

    Focus: land transport systems, public transport, and data-driven mobility research.

0x06 Related Repositories

1. paper with code

2. https://github.com/topics/traffic-prediction

3. Awesome-Trajectory-Prediction

4. traffic_prediction

5. transdim

   This project aim at provide accurate and efficient solution for spatio-temporal data prediction.

6. Urban Data Party

7. Multivariate Time Series Forecasting

8. deep-learning-time-series

9. GNN paper

10. Discovering millions of datasets on the web

11. LibCity An open-source research platform for intergrating several algorithms, data, and evalution metrics for traffic prediction.

12. GNN4Traffic

13. Traffic trajectory data/tools/papers collection

14. AGC-net

15. Forecasting on New Roads

16. ……

0x07 Benchmark & Evaluation

1. Task definition: Focus on short-term traffic flow/speed forecasting, typically evaluated with multi-horizon predictions (15/30/60 minutes).
2. Common datasets: METR-LA and PEMS-BAY are the most widely used public benchmarks. Recommend documenting time granularity, sampling interval, missing rate, and spatial coverage.
3. Common metrics: MAE, RMSE, MAPE. Recommend reporting metrics across multiple horizons, not a single step.
4. Splits: Use chronological train/val/test splits to avoid leakage from random splits.
5. Reproducibility: Provide random seeds, hardware details, training time, and parameter counts.

Benchmark Matrix

Use consistent metric definitions and masking policies when comparing papers.

Auto Trend Visuals

• Monthly candidates: updates/monthly-paper-candidates.md
• Word cloud and top terms: updates/wordcloud/
• Yearly hotspots and theme trends: updates/hotspots/
• Pipeline: .github/workflows/monthly-paper-search.yml (opens a PR for review when artifacts change)

Contribution

To make contributions on this repo, visit here