Discrimination analysis of EEG signals at eye open and eye close condition for ECS switching system

Dependable operation of brain computer interface (BCI) needs accurate classification of EEG. Application based on environment control system for disable people need comfortable and simple switching modes. This paper studies about eye close (EC) and eye open (EO) conditions of EEG for communication tool for severely disable users. For this purpose, Multivariate Gaussian Distribution analysis is applied to check the discrimination of two classes. Power Spectral Density and Central Tendency Measurement are used as features. These features used in Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA) and Fisher´s Linear Discriminant Analysis (FLDA) to observe the classification performance in linear and nonlinear environment. The experiment results show that EEG signals can be a reliable media for a switching paradigm of ECS.