EEG Emotion Recognition Based on Graph Regularized Sparse Linear Regression

Abstract

In this paper, a novel regression model, called graph regularized sparse linear regression (GRSLR), is proposed to deal with EEG emotion recognition problem. GRSLR extends the conventional linear regression method by imposing a graph regularization and a sparse regularization on the transform matrix of linear regression, such that it is able to simultaneously cope with sparse transform matrix learning while preserve the intrinsic manifold of the data samples. To detailed discuss the EEG emotion recognition, we collect a set of 14 subjects EEG emotion data and provide the experiment results on different features. To evaluate the proposed GRSLR model, we conduct experiments on the SEED database and RCLS database. The experimental results show that the proposed algorithm GRSLR is superior to the classic baselines. The RCLS database is made publicly available and other researchers could use it to test their own emotion recognition method.