PeMSD4: Traffic Flow Data for Spatiotemporal Prediction

A traffic flow prediction dataset used for evaluating the STF-GGRU model, which achieved an RMSE of 27.18 on this data. The dataset was uploaded by Karimeh Ibrahim Ata to figshare and last updated on 2026-04-28. It is stored in an XLS file with a size of 5.5 KB.

Use Cases

• Benchmarking traffic flow prediction models based on the reported RMSE performance metrics.
• Developing graph-based neural networks to capture non-Euclidean road network structures mentioned in the description.
• Training models to capture dynamic spatiotemporal interactions, as referenced by the study's focus on real-time adaptation.
• Comparing model performance against traditional methods like ARIMA, GRU, and LSTM, as cited in the description.

Strengths

• Dataset is associated with a specific model evaluation, reporting an RMSE of 27.18.
• Released under a permissive CC-BY-4.0 license.
• File size is precisely 5.5 KB, indicating a compact, focused dataset.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, uploaded by Karimeh Ibrahim Ata.

Collection Method

Likely collected from traffic sensors, as inferred from the domain.

Time Range

null

Freshness

Last updated 2026-04-28 17:37:42; freshness should be verified.

Geography

null