DIFFERENTIATING SKILLED AND UNSKILLED DRIVERS BY USING AN ADABOOST CLASSIFIER FOR DRIVER'S OPERATIONS

摘要

Prior to developing driving assistance systems to improve driving skills, the differences between skilled and unskilled drivers must be clarified. Previous studies suggest that highly skilled drivers share certain characteristics in accelerator, brake, and steering operations. This paper proposes a statistical method to extract these characteristics from data obtained in driving simulator experiments of driving around curves. The proposed method is composed of wavelet transform and an AdaBoost algorithm, a machine learning algorithm.