Mental Workload Classification for Multitasking Test using Electroencephalogram Signal

Mental Workload (MWL) can be explained as how much mental resource is required to perform any task. The measurement of mental workload can help to reduce mental fatigue, stress, anxiety, etc. MWL is playing a major role in the Education, Driving, Air traffic Control and Healthcare system. In this paper, several time-domain features are calculated on the dataset to classify the workload. The workload is classified as low mental workload or high mental workload based on the rating after completion of the mental task by different subjects. The dataset consists of Electroencephalogram (EEG) signals that are recorded during the mental task performed by subjects. EEG data were recorded from 14 channels by emotive epoch device. Support Vector Machine (SVM) with Radial Basis Function (RBF), Linear, Sigmoid and Poly kernels is used to classify the levels of mental workload. SVM is applied on each feature to calculate the performance measure. The mental workload is classified with 88.88 % accuracy by SVM with linear kernel.