Abstract: RTAs are a major global health issue, resulting in significant injury, death, and economic costs. The goal of this research is to develop a prediction model to assess the risk associated with various accident scenarios by investigating the numerous elements that impact the severity of RTAs. The study analyzes data from police-reported traffic accidents over a five-year period, considering variables such as driver demographics, vehicle type, environmental conditions, and road characteristics. The findings indicate that accident severity is strongly affected by factors like the age and gender of the driver, time of day, weather conditions, and the type of road. Accidents involving young drivers, male drivers, and those occurring at night or in bad weather are more severe. Additionally, crashes on RR and those involving heavy vehicles tend to be more serious. To determine the probability of a serious injury or death in a rear-end collision, a predictive model employing ML and LR approaches was created. This model, which showed high accuracy, can help policymakers and traffic safety officials identify high-risk situations and implement specific preventive measures. The study highlights the need for comprehensive road safety strategies that include enforcement, education, and engineering improvements.

Keywords: RTAs-Road Traffic Accidents, ML-Machine Learning, LR-Logistic Regression, RR-Rural Roads.

| DOI: 10.17148/IARJSET.2024.11765