Data: Characterisation of Cognitive Load using Machine Learning Classifiers of Electroencephalogram Data
A high cognitive load can overload a person resulting in catastrophic accidents, it is therefore important to ensure the cognitive load of a safety-critical task (such as driving a vehicle) is at a manageable level. Although electroencephalography (EEG) has attracted significant interest in the research of cognitive load, few studies use EEG to investigate driving-related cognitive load. This paper presents a feasibility study on the simulation of various levels of cognitive load through designing and implementing four driving tasks, and the associated classification of load using EEG recordings. An EEG dataset containing these four driving tasks from a group of 20 participants was collected to investigate whether EEG can be used as a biomarker to reflect changes in cognitive load. The dataset was then used to train four Deep Neural Networks (DNNs) and four Support Vector Machines (SVMs) classification models. The results showed that the best model achieved a classification accuracy of 90.37% utilizing statistical features from multiple frequency bands in 24 EEG channels. Furthermore, it was observed that the Gamma and Beta bands achieved a greater classification accuracy than the Alpha and Theta bands. The output of this study can potentially improve the Human-Machine-Interface of vehicles for enhanced safety.