Time series and support vector machines to predict Powered-Two-Wheeler accident involvement and accident type, 2017

A paper titled “Time series and support vector machines to predict Powered-Two-Wheeler accident involvement and accident type” co-authored by Athanasios Theofilatos, George Yannis, Costas Antoniou, Antonis Chaziris and Dimitris Sermpis, is now published in Journal of Transportation Safety and Security. This study exploited real-time traffic and weather data from two major urban arterials in the city of Athens, Greece. Due to the high number of candidate variables, a random forest model was applied to reveal the most important variables. Then, the potentially significant variables were used as input to a Bayesian logistic regression model in order to reveal the magnitude of their effect on PTW accident involvement. The results of the analysis suggest that PTWs are more likely to be involved in multi-vehicle accidents than in single-vehicle accidents. It was also indicated that increased traffic flow and variations in speed have a significant influence on PTW accident involvement.  ΠΗΓΗ: Ε.Μ.Π    www.nrso.ntua.gr