Detecting anomalies in traffic data using a flexible semi-parametric model
Abstract The paper addresses the task of improving the quality of traffic data by combining semantic analysis and statistical modeling. The study focuses on dynamic traffic variables, specifically speed and volume of traffic flow. The results are presented based on semantic quality control outputs f...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-05-01
|
| Series: | European Transport Research Review |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12544-025-00728-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract The paper addresses the task of improving the quality of traffic data by combining semantic analysis and statistical modeling. The study focuses on dynamic traffic variables, specifically speed and volume of traffic flow. The results are presented based on semantic quality control outputs for dynamic traffic variables. The semantic analysis aimed to find and label segments of time series waveforms that fell outside the expected value range at specific locations and times. This approach allowed for a more detailed analysis of specific cases, enhancing our understanding of traffic dynamics. We employed a Generalized Additive Model (GAM) to detect non-standard data segments successfully. Real traffic data from the motorway in the Czech Republic verified the effectiveness of the chosen statistical approach. |
|---|---|
| ISSN: | 1866-8887 |