Chapter 5: Using machine learning and deep learning for traffic congestion prediction: a review
Traffic congestion has long been a problem for many cities and commuters around the world, which causes long commuting hours, increases traffic crash rates and results in significant economic and productivity losses. Correctly predicting traffic congestion can help alleviate several problems that tr...
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Online |
| Sprache: | Englisch |
| Veröffentlicht: |
Edward Elgar Publishing
2026
|
| Schlagworte: | |
| Online-Zugang: | https://directory.doabooks.org/handle/20.500.12854/173417 |
| Tags: |
Keine Tags, Fügen Sie das erste Tag hinzu!
|
| Zusammenfassung: | Traffic congestion has long been a problem for many cities and commuters around the world, which causes long commuting hours, increases traffic crash rates and results in significant economic and productivity losses. Correctly predicting traffic congestion can help alleviate several problems that traffic congestion causes on a recurrent basis. With the advances in data collection, artificial intelligence (AI) becomes an ideal tool for short-term and long-term congestion forecasting. This chapter reviews the latest developments in machine learning and deep learning methodologies for traffic congestion prediction in a systematic way, covering literature over the last decade. The main findings are structured based on different AI methodologies, datasets and prediction time periods. The chapter also discusses the advantages and drawbacks of current AI methodologies and describes the research gaps that must be overcome to enable real-world implementation of AI methodologies for traffic congestion prediction. |
|---|