Improving Strategic Policies for Pedestrian Safety Enhancement Using Classification Tree Modeling
- Title
- Improving Strategic Policies for Pedestrian Safety Enhancement Using Classification Tree Modeling
- Author
- 정소영
- Keywords
- Pedestrian safety enhancement; Injury severities; Classification tree; Contributing factors; Strategic policies; TRAFFIC INJURY SEVERITY; NEW-YORK-CITY; VEHICLE CRASHES; CLASSIFIERS; MULTICLASS; REGRESSION; IMPACT
- Issue Date
- 2016-01
- Publisher
- PERGAMON-ELSEVIER SCIENCE LTD
- Citation
- TRANSPORTATION RESEARCH PART A-POLICY AND PRACTICE, v. 85, Page. 53-64
- Abstract
- Pedestrian safety enhancement is a key component in reducing traffic fatalities in the Republic of Korea. The purpose of this study was to review, validate, specify, and prioritize Korea’s strategic policies for pedestrian safety enhancement using the classification tree method to model pedestrian injury severities. The findings show that pedestrian age and movement type are the two primary variables contributing to pedestrian fatalities and severe injuries. Traffic operation, road class, crash location, driver violation, and at-fault vehicle type are all secondary variables associated with pedestrian fatalities and severe injuries. Factors that contributed to crashes were compared with strategic polices for senior zones and school zones, road safety facilities, safe walking environments, and legal obligations of the driver in order to understand why certain polices are ineffective versus effective. Consequently, this study provides prescriptive analysis and specific insights pertaining to strategic policies for pedestrian safety enhancements, which can be employed in other countries for the similar purpose. For further research, this study suggests combining several other data-mining techniques with nationwide data collection.
- URI
- https://www.sciencedirect.com/science/article/pii/S0965856416000021http://hdl.handle.net/20.500.11754/48525
- ISSN
- 0965-8564
- DOI
- 10.1016/j.tra.2016.01.002
- Appears in Collections:
- RESEARCH INSTITUTE[E](부설연구소) > ETC
- Files in This Item:
There are no files associated with this item.
- Export
- RIS (EndNote)
- XLS (Excel)
- XML