Evaluation of Contributing Factors Affecting Number of Vehicles Involved in Crashes Using Machine Learning Techniques in Rural Roads of Cosenza, Italy

The evaluation of road safety is a critical issue having to be conducted for successful management in transport systems, whereas considered transportation systems as challenging task according the dynamic this and presence large number effective parameters on safety. Therefore, analysis important contributing factors affecting vehicles involved crashes play key role increasing efficiency For purpose, research work, two machine learning algorithms, including group method data handling (GMDH)-type neural network combination support vector (SVM) grasshopper optimization algorithm (GOA), are employed. Hence, an accident output, seven safety, Daylight (DL), Weekday (W), Type (TA), Location (L), Speed limit (SL), Average speed (AS), Annual average daily traffic (AADT) rural roads Cosenza, southern Italy, selected inputs. In study, 564 sets from areas were investigated, relevant, measured. next stage, several models developed investigate areas. results obtained demonstrated that “Type accident” has highest level “Location” lowest importance investigated area. Finally, although both algorithms same, GOA-SVM model showed better degree accuracy robustness than GMDH model.