IoT Vehicle Traffic Anomaly Detection Using Machine Learning Techniques: A Case Study with DAD Dataset

Abstract

The increasing growth of the Internet of Things (IoT) has significantly transformed various sectors, including transportation. With the increasing number of connected vehicles, the need for effective traffic anomaly detection systems has become crucial for ensuring safety and efficiency on the roads. This study aims to address this need by employing Machine Learning (ML) techniques to detect anomalies in vehicle traffic, specifically using the Driver Anomaly Detection (DAD) dataset. The ML methods employed in this study include Logistic Regression (LR), Support Vector Machine (SVM), and Random Forest (RF), with RF showing the best performance with an accuracy of 94.27%. The RF model also exhibited the highest Area Under the Curve (AUC) value of 0.9813, outperforming existing studies. These findings highlight the potential of RF for effectively identifying traffic anomalies. Based on these results, future research could explore deep learning techniques and edge computing to further enhance detection accuracy and system effectiveness. Additionally, expanding the study to diverse datasets and investigating hybrid models might improve anomaly detection systems.