Applying Machine Learning for Improving Performance Classification on Driving Behavior

Abstract

Driving accidents are serious events that could cause fatality. According to WHO’s reports, reckless driving behaviors such as speeding, driving under influence, and operating phones while driving are among the main factors that could reduce the focus of drivers while driving. Driving accidents are also difficult to handle on a large scale in a country. Using machine learning classification method to determine whether a driver is driving safely or not can help reduce the risk of driving accidents. Drivers tend to be more careful when they know that their driving behaviors are being monitored. We created a classifier model that can be applied to detection systems to classify whether a driver is driving safely or not safely using travel sensor data, which includes gyroscope, accelerometer, and GPS. The classification methods used in this study are Random Forest (RF) classification algorithm, Support Vector Machine (SVM), and Multilayer Perceptron (MLP). This study shows that RF has the best performance with 98% accuracy, 98% precision, and sensitivity 97%. Performance testing shows that the proposed pre-processing method can increase the classifier sensitivity value in the research dataset. It is hoped that the classifier model can be applied to the driving detection system so that it can reduce the risk of traffic accidents.