Traffic injury prevention · 2026
Machine learning and Geographic Information Systems (GIS)-based model for road crash severity prediction and behavioural pattern analysis.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
A Al-Mashagba Lara, Sumari Putra, Mashagba Mohammadnour, Al Abri Fathiya, Mashagba Hamza A, Abd Aziz Azlan B, Al-Bawri Samir Salem
Аннотация
OBJECTIVES: Road traffic injuries constitute a major public health problem in low- and middle-income countries, including Jordan. This study develops a hybrid machine learning and Geographic Information Systems (GIS)-based framework to predict crash severity and identify behavioral, environmental, and infrastructural factors contributing to injury outcomes.
METHODS: A dataset of 11,345 crashes from the Jordan Traffic Institute (2018) was analyzed. After encoding, imputation, normalization, and outlier treatment, a two-stage analytical design was applied. First, association rule mining (Apriori; minimum support 0.05, confidence 0.60) was used to uncover dominant behavioral and environmental patterns linked to different injury levels. Second, Decision Tree, Random Forest, and AdaBoost models were trained (70/30 split) to classify severity (slight, medium, severe, fatal). Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrices. GIS-based kernel density estimation was used to detect spatial hotspots of severe and fatal crashes.
RESULTS: Association rules revealed strong links between nighttime driving, young drivers, and lack of seatbelt use with fatal outcomes. GIS analysis identified high-risk clusters on major urban area in Zarqa, Awajan, and Al-Rusayfah. The Random Forest model achieved the highest predictive performance (accuracy: 98.5-99.9%). Key predictors included crash type, vehicle speed, time of day, driver age, and road characteristics.
CONCLUSIONS: The proposed hybrid ML-GIS framework offers highly accurate severity prediction and reveals spatial and behavioral patterns critical for targeted safety interventions. Findings highlight the dominant influence of temporal and infrastructural factors over driver characteristics, supporting evidence-based engineering and enforcement strategies in Jordan.
METHODS: A dataset of 11,345 crashes from the Jordan Traffic Institute (2018) was analyzed. After encoding, imputation, normalization, and outlier treatment, a two-stage analytical design was applied. First, association rule mining (Apriori; minimum support 0.05, confidence 0.60) was used to uncover dominant behavioral and environmental patterns linked to different injury levels. Second, Decision Tree, Random Forest, and AdaBoost models were trained (70/30 split) to classify severity (slight, medium, severe, fatal). Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrices. GIS-based kernel density estimation was used to detect spatial hotspots of severe and fatal crashes.
RESULTS: Association rules revealed strong links between nighttime driving, young drivers, and lack of seatbelt use with fatal outcomes. GIS analysis identified high-risk clusters on major urban area in Zarqa, Awajan, and Al-Rusayfah. The Random Forest model achieved the highest predictive performance (accuracy: 98.5-99.9%). Key predictors included crash type, vehicle speed, time of day, driver age, and road characteristics.
CONCLUSIONS: The proposed hybrid ML-GIS framework offers highly accurate severity prediction and reveals spatial and behavioral patterns critical for targeted safety interventions. Findings highlight the dominant influence of temporal and infrastructural factors over driver characteristics, supporting evidence-based engineering and enforcement strategies in Jordan.