Scene Identification of Single-Motorcycle Death Accidents Based on Binary Logistic Regression and Association Rules

The mortality of single-motorcycle accidents is significantly higher than other types of crashes involving motorcycles. In order to prevent the occurrence of single-motorcycle deaths, accident scenes are identified from three aspects: drivers, roads, and environment. Firstly, the factors influencing the severity of single-motorcycle crash accidents are researched. Collecting 4,067 single-motorcycle crash items with 13 independent variables involving drivers’ characteristics, road conditions, and the environment, a binary logistic model is constructed to account for the correlation between the 13 factors and severity of accidents. Secondly, using the association rules based on Apriori algorithm, 10 different death scenes involving single-motorcycle and considering characteristics of motorcyclists, roads, and environments are identified. At the same time, a series of death prevention measures are proposed for the traffic management departments. The method of identifying death scenes put forward in this paper can be used as a theoretical basis for the prevention of similar traffic accidents.