DEEP LEARNING-BASED ASSESSMENT MODEL FOR IDENTIFICATION OF VISUAL LEARNING STYLE USING RAW EEG SIGNALS

Learning style has its importance especially for long-term learning provided that an appropriate style is selected. The importance of determining a suitable learning style using brain patterns cannot be ignored as suggesting learning style without knowing brain patterns can increase the cognitive load. In the literature, various studies based on electroencephalography (EEG) have been proposed to identify the learning style. However, the utility of these methods is not clear as they lack a common framework. Also, as these methods are based on self-assessment, they give biased results that warrant further research. The objective of this study was to develop an EEG based assessment model for the identification of visual learning style. EEG signals were recorded during resting state (eye open, eye close) conditions and during performing learning tasks and recall tasks. Correct responses were analyzed for two recall sessions: Recall session
one and Recall session two. The EEG features, Power Spectral Density (PSD) and Discrete Wavelet Transform (DWT) feature extracted from frontal, occipital and parietal brain regions
were found to be the most significant for identifying the visual learning styles of students. The feature selection is done using principal component analysis (PCA).