Prediction of Road Accident Severity Using Machine Learning Algorithm
Injuries due to road accidents are one of the most prevalent causes of death apart from health related issues. The World Health Organization states that road traffic injuries caused an estimated 1.35 million deaths worldwide in the year 2016. That is, a person is killed every 25 seconds. This calls for the need to analyse road accidents and the factors affecting them and come up with a method to reduce the probability of their occurence. The analysis of road accident severity was done by running an accident dataset through several machine learning classification algorithms to see which model performed the best in classifying the accidents into severity classes such as slight, severe and fatal.. It was observed that logistic regression to perform multilabel classification gave the highest accuracy score. It was also observed that factors such as number of vehicles, lighting conditions and road features played a role in determining the severity of the accident.