Abstract

• In the case of road accidents, millions of people are dead each year, and the numbers of accident rates are rising all over the world. As a result, they have a great impact on society in terms of finance and economic. In this paper, Machine Learning (ML) models are used to analyze road accident severity based on road accident dataset. The dataset was collected from Myanmar’s Traffic Police Force which shares raw data on annual basis on Yangon- Naypyitaw expressway’s traffic accident data in 2022. The dataset is divided into training and testing dataset. The training dataset is used to train our model, and the test dataset is utilized to evaluate the predictions. In the proposed system, the collected data is preprocessed (data cleaning, encoding, transformation) and followed by data training, testing and comparison analysis on the analysis of the ML methods. By this experiment, the road accident contributing factors is vital. By using the continuous variables, this work is predicted road accidents with different accident types and applied with ML technique like Logistic Regression (LR), Adaptive Boosting (AdaBoost), Decision Tree, Adaboost using Decision Tree and Multinomial Naive Bayes. The experimental results showed that Logistic Regression classifier achieves the best accuracy than other classifiers.