Abstract
Road accidents caused by weather conditions in winter lead to higher mortality rates than in other seasons. The main causes of road accidents include human carelessness, vehicle defects, road conditions, and weather factors. If the risk of road accidents with changes in road weather conditions can be quantitatively evaluated, it will contribute to reducing the road accident fatalities. The road accident data used in this study were obtained for the period 2017 to 2019. Spatial interpolation estimated the weather information; geographic information system (GIS) and Shuttle Radar Topography Mission (SRTM) data identified road geometry and accident area altitude; synthetic minority oversampling technique (SMOTE) addressed the data imbalance problem between road accidents due to weather conditions and from other causes, and finally, machine learning was performed on the data using various models such as random forest, XGBoost, neural network, and logistic regression. The training- to test data ratio was 7:3. Random forest model exhibited the best classification performance for road accident status according to weather risks. Thus, by applying weather data and road geometry to machine learning models, the risk of road accidents due to weather conditions in the winter season can be predicted and provided as a service.
Highlights
The ever-increasing vehicular traffic has resulted in corresponding increase in fatalities due to traffic accidents [1]
Shuttle Radar Topography Mission (SRTM) data identified road geometry and accident area altitude; synthetic minority oversampling technique (SMOTE) addressed the data imbalance problem between road accidents due to weather conditions and from other causes, and machine learning was performed on the data using various models such as random forest, XGBoost, neural network, and logistic regression
We aim to investigate the causes of past traffic accidents on highways in winter and examine the relationship between factors affecting accidents, such as road weather conditions and road geometry using various machine learning models
Summary
The ever-increasing vehicular traffic has resulted in corresponding increase in fatalities due to traffic accidents [1]. During the winter season in the Korean peninsula (with four distinct seasons), the roads are covered with a large amount of ice. During the winter season in the Korean peninsula (with four distinct seasons), the roads are covered with a large amount of ice In this regard, a service that presents prediction and guidance on the risks of traffic accidents due to weather conditions in winter may reduce the road accidents fatalities. Changes in snowfall, rainfall, and weather have a significant impact on road safety, by reducing the driver’s visibility, and the friction between the vehicle and the road [4]. Among these causes, factors such as human carelessness and vehicle defects occur by chance, making it difficult to predict and provide guidance in advance. We aim to investigate the causes of past traffic accidents on highways in winter and examine the relationship between factors affecting accidents, such as road weather conditions and road geometry using various machine learning models
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
