A source of traffic sequence data and auxiliary information used for the KDD 2018 paper on deep sequence learning for traffic prediction. It contains time-series traffic observations paired with external factors to support multi-step forecasting tasks in urban environments.
Use Cases
- Train deep sequence models for traffic prediction using the provided traffic sequence and auxiliary feature sets.
- Benchmark time-series forecasting algorithms against the results published in the KDD 2018 paper.
- Analyze the correlation between auxiliary information and traffic flow patterns to improve urban mobility planning.
Strengths
- Features traffic sequence data for urban road networks as used in the KDD 2018 research.
- Includes auxiliary information categories designed to improve deep sequence learning model accuracy.
- Provides the original benchmark data and code for reproducing traffic prediction results.