A comprehensive review of various data collection approaches, features, and algorithms used for the classification of driving style

Abstract

In the transport sector problems, road safety is a prime concern in emerging nations. Applications about driving assistance are being actively studied to address road safety matters, including humanistic performance defined as one of the principal causes for problems on road safety, which confirms why driving style is currently experiencing extensive research attention. Future driving style prediction will form the basis for eco-driving and energy management strategies. From this aspect, analyzing drivers' behavior is necessary to improve road safety. The comprehensive survey provides a summary, outline, and structure a large collection of work from various sources and proposes a comparison of the devices used, parameters studied, and classification algorithms used for analyzing driving style. The researches done on the driving style analysis uses a wide variety of devices and several parameters. This analysis shows that a diverse set of parameters that can be used to analyze the style of driving and seeks to understand the various machine learning classification methods and metrics for the classification of driving style.