Scholarlink

The rising rate of road traffic accidents, due to increased travel demand, has raised concerns among transport experts in countries across the Global South. This study analyzed the characteristics of road traffic accidents in the region using machine learning algorithms - Multinomial Logistic Regression (MLR), Naïve Bayes (NB), Support Vector Machine (SVM), and Random Forest (RF), focusing on the case of Makurdi Metropolis in Nigeria. Objectives of the study included identifying key factors influencing motorists’ involvement in road accidents, understanding motorists’ perceptions regarding accident causes, and recommending policy strategies for sustainable road safety. A sample size of 400 was selected from a population of 10,000 registered motorists in Makurdi Metropolis. Data were collected using a self-structured and closed-ended questionnaire administered face-to-face through a random sampling technique. Results of the study indicated that factors influencing road accidents include driving experience, gender, level of education, status of driving license, and driving frequency. Also, the MLR, NB, and SVM models achieved crash severity level classification accuracies of 52.90%, 52.20%, and 49.90% respectively. The RF model outperformed others with an accuracy of 73.40%. The strict enforcement of traffic regulations, public awareness of road safety rules, and the installation of traffic signals were strongly recommended as the major causes of road traffic accidents. These strategies are essential for minimizing road traffic accidents to promote sustainable transport development in the Global South.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Road Traffic Accident Analysis Using Machine Learning Techniques: A Context of an African City

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • P. T. Adeke,
  • D. K. Das,
  • M. M. H. Mostafa,
  • P. A. Ochayi

摘要

The rising rate of road traffic accidents, due to increased travel demand, has raised concerns among transport experts in countries across the Global South. This study analyzed the characteristics of road traffic accidents in the region using machine learning algorithms - Multinomial Logistic Regression (MLR), Naïve Bayes (NB), Support Vector Machine (SVM), and Random Forest (RF), focusing on the case of Makurdi Metropolis in Nigeria. Objectives of the study included identifying key factors influencing motorists’ involvement in road accidents, understanding motorists’ perceptions regarding accident causes, and recommending policy strategies for sustainable road safety. A sample size of 400 was selected from a population of 10,000 registered motorists in Makurdi Metropolis. Data were collected using a self-structured and closed-ended questionnaire administered face-to-face through a random sampling technique. Results of the study indicated that factors influencing road accidents include driving experience, gender, level of education, status of driving license, and driving frequency. Also, the MLR, NB, and SVM models achieved crash severity level classification accuracies of 52.90%, 52.20%, and 49.90% respectively. The RF model outperformed others with an accuracy of 73.40%. The strict enforcement of traffic regulations, public awareness of road safety rules, and the installation of traffic signals were strongly recommended as the major causes of road traffic accidents. These strategies are essential for minimizing road traffic accidents to promote sustainable transport development in the Global South.