Predicting the type of road accidents based on air temperature in Iran: A case study of roads in Qazvin province

Abstract

This study investigates the relationship between ambient temperature, weather conditions, and types of road accidents in Qazvin province, Iran. The research addresses a significant societal challenge of road accidents, particularly in developing countries like Iran. The objectives are to analyze the correlation between temperature and accident types and to develop a predictive model using data mining techniques. The study employs a quantitative approach, analyzing over 15,000 accident records from 2010 to 2020. The findings reveal a connection between the temperature variable and the type of road accidents as well as weather conditions. Additionally, data mining analysis identifies a predictable pattern among temperature variables, types of road accidents, and weather conditions. Implications of the study underscore the importance of considering temperature and weather conditions as secondary factors influencing accidents. The predictive model can aid decision-makers in formulating effective strategies to reduce accidents. Understanding the relationship between temperature, weather, and accident types enables the design of targeted interventions to enhance road safety. This research contributes valuable insights to accident reduction efforts and emphasizes the significance of addressing environmental variables in road safety planning and policy-making. Moreover, the results of the data mining pattern analysis indicate that car overturning accidents in various weather conditions are the primary type of accidents, followed by chain accidents. However, the types of accidents vary based on different weather conditions and temperatures. The study highlights the intricate connection between weather conditions, temperature, and types of road accidents. By utilizing data mining techniques, the research provides a predictive model for accident patterns, offering valuable insights to enhance road safety strategies.

Keywords: Data mining; Decision making; Road accident; Temperature; Weather condition.

Fig. 1

Modeling (decision tree) of road accidents based on the type and manner of occurrence of road accidents in Qazvin province with temperature variable.