|
|
|
AUTHOR(S): Mohammed Akour, Osama Al Qasem, Firas Hanandeh
|
|
TITLE |
 PDF |
|
ABSTRACT Based on several reports, one of the main causes of human injuries and death is Traffic accidents. Many Communities are suffering from the accidents at different level of severity. Traffic accident severity prediction might play a role in enhancing the management and controlling the safety traffic. By utilizing existing road accident data, more accuracy of accident severity prediction can be performed. This research paper aims to build an accurate traffic accident severity prediction model. The proposed model is mainly based on ensemble machine learning algorithms i.e. Random Forest, XGBoost and decision tree. For comparison purposes, the performance of the studied ensemble methods is compared with the base learners. Four measurements are recorded and used for comparison. The findings of this papers shows that Balanced Random Forest, XGBoost and decision tree provide a promising tool for predicting the injury severity of traffic accidents. Moreover, the voting (hard) has an advantage over the other two representative classifiers. Compared with other classifiers, voting (hard) has a good ability to predict fatal/serious injury. |
|
KEYWORDS Traffic Accident Severity prediction, Injuries, Base learner Algorithms, Ensemble Algorithms |
|
|
|
Cite this paper Mohammed Akour, Osama Al Qasem, Firas Hanandeh. (2022) Traffic Accident Severity Prediction: A comparison Study. International Journal of Transportation Systems, 7, 24-28 |
|
|
