Older Driver Crash Risk Modeling by AdaBoost

Accompanying the progression of aging population, there has been a considerable change in the distribution of licensed drivers in different age groups, which led to significant complications in traffic safety. Noticeably, older drivers were overrepresented in a variety of types of crashes. In this paper, the authors set out to investigate the characteristics of crash risk for older drivers and based the study on two years of statewide crash data in California from 2010 to 2012. By chosen to adopt a machine learning algorithm, called Adaptive Boost (AdaBoost), and developed a model to classify characteristic of crash risk describing the relationship between crash risk and twelve variables recorded in the crash database. The methodology was selected to overcome the linearity limitation of regression models that are typically used for this type of analysis. The difference of crash characteristics were presented for older drivers aged 65 and higher, with a comparison with young drivers aged from 20 to 39. The finding shows that older drivers are more easily involved in crashes due to weather and road surface factors. Additionally, older drivers have higher risks when driving at highways and intersections, especially for U-turns. Because of functional limitations of older drivers such as vision and attention, older drivers also have higher crash probabilities in head-on collisions. The outcome of this study shows that the AdaBoost model can be useful in identifying the higher risk conditions for older drivers and planning accordingly to mitigate safety concerns.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ANB60 Safe Mobility of Older Persons.
• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Du, Yaoqiong
  • Chan, Ching-Yao
• Conference:
  • Transportation Research Board 93rd Annual Meeting
  • Location: Washington DC
  • Date: 2014-1-12 to 2014-1-16
• Date: 2014

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 18p
• Monograph Title: TRB 93rd Annual Meeting Compendium of Papers

Subject/Index Terms

• TRT Terms: Aged drivers; Crash characteristics; Crash data; Crash risk forecasting; Human factors in crashes
• Geographic Terms: California
• Subject Areas: Data and Information Technology; Highways; Safety and Human Factors; I81: Accident Statistics; I83: Accidents and the Human Factor;

Filing Info

• Accession Number: 01518627
• Record Type: Publication
• Report/Paper Numbers: 14-3532
• Files: TRIS, TRB, ATRI
• Created Date: Mar 20 2014 1:39PM