A Stacking Ensemble Learning Model for Mental State Recognition towards Implementation of Brain Computer Interface

This paper presents a novel stacking ensemble learning model that aims at improving mental state classification for brain computer interface implementation. The proposed model combines machine learning based methods that use support vector machine, artificial neural network and deep learning with a model selection rule to classify EEG signals into accurate mental states. The proposed ensemble learning model is validated on an EEG dataset in which EEG signals are recorded from four subjects. Three mental tasks are turning mind into Zen condition (Neutral), imagining how to turn a light on (Light) and memorizing scientific paper content (Paper). Experimental results show that this ensemble learning model is robust and effective while comparing with other existing machine learning models and methods for the same purpose.