Klasifikasi Keparahan Kecelakaan Lalu Lintas Menggunakan Machine Learning

Abstract

Traffic accidents are one of the main causes of injury and death in various countries. Various factors such as the number of vehicles involved, lighting conditions, type of intersection, and the underlying cause of the accident contribute to the severity of the accident. This research aims to classify the severity of traffic accidents using Random Forest, Decision Tree, Naïve Bayes, and Support Vector Machine (SVM) algorithms. The dataset used was obtained from Kaggle with various relevant categorical and numerical features. The research methodology includes data collection and model evaluation. Model evaluation is carried out using accuracy, precision, recall, and F1-score. The research results show that Random Forest provides the highest accuracy of 84%, followed by Naïve Bayes (82%), Decision Tree (77%), and SVM (71%). The SMOTE technique improves the performance of some models by improving data balance, but also increases the risk of overfitting. Meanwhile, applying PCA helps improve model efficiency by reducing dimensions without losing important information. With the model that has been developed, this prediction system has the potential to be used in road safety analysis and to support decision making in mitigating traffic accidents