Emotion detection from EEG recordings based on supervised and unsupervised dimension reduction

摘要

In recent years, researchers have been trying to detect human emotions from recorded brain signalssuch as electroencephalogram (EEG) signals. However, due to the high levels of noise fromthe EEG recordings, a single feature alone cannot achieve good performance. A combination ofdistinct features is the key for automatic emotion detection. In this paper, we present a hybriddimension feature reduction scheme using a total of 14 different features extracted from EEGrecordings. The scheme combines these distinct features in the feature space using both supervisedand unsupervised feature selection processes. Maximum Relevance Minimum Redundancy(mRMR) is applied to re-order the combined features into max-relevance with the labels andmin-redundancy of each feature. The generated features are further reduced with principal componentanalysis (PCA) for extracting the principal components. Experimental results show thatthe proposed work outperforms the state-of-art methods using the same settings in the publiclyavailable DEAP data set.