A META-FRAMEWORK USING ENSEMBLES FOR EEG DIAGNOSIS

Abstract

It enables people to communicate with computers by using their brains. Electroencephalography (EEG) data are often used to quantify this sort of activity. A general time series problem for recognizing human cognitive states is eye state classification. Knowing human cognitive states can be quite useful for therapeutic applications in our daily life. Analyses that are both subject-dependent and independent are used to classify the current ocular states. In subject-dependent classification, the model is trained using data from a subject. Subject-specific categorization, however, is exempt from this requirement. There are issues with the EEG data because of noise and muscle activity. This study suggested a categorization approach that employs a separate pre-processing stage. In this context, the basis classifiers and the most significant studies are compared to the ensemble techniques used in the classification step. A publicly accessible EEG eye state dataset from UCI is used for evaluation. The results are 96.99%.