Search

Machine learning (ML) techniques have higher prediction accuracy compared to conventional statistical methods for crash frequency modelling. However, their black-box nature limits the interpretability. The objective of this research is to combine both ML and statistical methods to develop hybrid link-level crash frequency models with high predictability and interpretability. For this purpose, M5' model trees method (M5') is introduced and applied to classify the crash data and then calibrate a model for each homogenous class. The data for 1134 and 345 randomly selected links on urban arterials in the city of Charlotte, North Carolina was used to develop and validate models, respectively. The outputs from the hybrid approach are compared with the outputs from cluster-based negative binomial regression (NBR) and general NBR models. Findings indicate that M5' has high predictability and is very reliable to interpret the role of different attributes on crash frequency compared to other developed models.

Objective: The time left to cross the street displayed on pedestrian countdown signals can be used by pedestrians as well as drivers of vehicles, though these signals are primarily provided to help pedestrians make better crossing decisions at signalized intersections. This article presents an evaluation of the effect of pedestrian countdown signals in reducing vehicle-pedestrian crashes and all crashes at signalized intersections., Methods: A before-and-after study approach was adopted to evaluate the effect considering pedestrian countdown signals installed over a 5-month period at 106 signalized intersections in the city of Charlotte, North Carolina., Results: Analysis conducted at 95 percent confidence level showed that there has been a statistically insignificant decrease in vehicle-pedestrian crashes but a statistically significant decrease in all (includes vehicle-pedestrian and vehicle(s) only involved) crashes after the installation of pedestrian countdown signals. No negative consequences were observed after the installation of pedestrian countdown signals. Sixty-eight percent of the signalized intersections saw a decrease in the total number of all crashes, and 4 percent of the signalized intersections have not seen any change in the number of all crashes after the installation of pedestrian countdown signals. Improvements in terms of decrease in the total number of all crashes was high at signalized intersections with greater than 10 crashes per year during the before period. Likewise, decrease in the number of all crashes was high at signalized intersections with traffic volume between 7 AM to 7 PM greater than 20,000 vehicles during the before period., Conclusions: Based on results obtained, it can be concluded that pedestrians as well as drivers are making better decisions using the time left to cross the street displayed on pedestrian countdown signals at signalized intersections in the city of Charlotte, North Carolina.