Driver behavior study for better speed prediction: unique use of speed sensor and machine learning

With the advent of artificial intelligence and data, intelligent vehicles have become a major topic both in academia and industry for the promises they exhibit. In general, driver behavior analysis is based through manufacturer information, this information is confidential or encrypted. Through this paper, we will explore the possibility of building a system that allows a personalized driving experience. For that, we use only the speed of the vehicle that can be recovered through external systems to the vehicle. After collecting non-invasive minimal sensing information, the driver can be identified through his speed where his own driving style is characterized and evaluated using both supervised and unsupervised machine learning approaches. With the same data, the prediction of the behavior of drivers on future journeys has been developed using the Markov method.