The Application of Hidden Markov Model in Classifying Novice and Experienced Drivers by Driving Behavioral Features

Experience is a kind of knowledge that people can get from everyday practice, which can make a novice an expert. Behavioral data showed that experienced driver and novice had different driving behavior patterns, and there existed a close relation between driving behavior pattern and driving performance. In this study, a simulated driving system was used as a driving environment. Based on 10 drivers' (including 5 drivers with more than 30 thousands kilometers driving and 5 just-got-license drivers) driving performance, a Hidden Markov Model (HMM) was built to find the critical features that differentiate expert driver and novice. The classification HMM model was verified by another 10 experienced + 10 novice drivers' driving performance. The correct recognition rate of the differentiation HMM model is more than 85%.