Predictive Analysis of Jordan Road Accidents Severity Using Machine Learning

Jordan has a high and growing level of traffic accidents reaching 160,600) accidents in 2021, (11,241) of them had human injuries, (589) deaths, and (320) million JOD losses. Road traffic accidents are currently the 8th leading cause of death worldwide, accounting for about 1.35 million fatalities annually. Understanding the primary causes of these accidents and the circumstances in which they occur is essential if governments throughout the world are to put policies in place to limit the number of fatalities brought on by traffic accidents. This project has two objectives: (i) to learn more about Jordan's present road accident situation by performing an exploratory study on several road accident datasets, and (ii) to investigate the effectiveness of various machine-learning techniques in predicting the severity of road accidents in Jordan. The dataset entities were collected, collated, explored, and prepared for use in the model. To assess their predictive performance, five different classification algorithms were trained and tested. The findings of the present study demonstrate that the best algorithm was Logistic Regression, which had an accuracy of 98.1%.