Understanding driver behavior using multi-dimensional CMAC

In this paper, detailed studies have been conducted to model individual driving behavior in order to identify features that may be efficiently and effectively used to profile each driver. The brake and gas pedal pressure are used to identify uniqueness in driving maneuver of each driver. These differences in the driving habits could be due to the way our subconscious mind works and respond. In addition the switching between the subconscious to conscious mind will also produce unique respond on how the brain perform. Since the activation of movements are controlled by the cerebellum we propose the use of cerebellum model articulation controller (CMAC), introduced by Albus, to model each driver behavior. In this paper we only focus on using the gas pedal and brake pedal pressure of the driver to understand the driver behavior under difference environment. Experimental results from the CMAC profiles show the potential of extracting features of drivers' behavior for identification, verification, emotion recognition, stress and many other behavioral conditions.