Loading...

Emotion Recognition from EEG Signals using Tensor based Algorithms

Einizadeh, Aref | 2018

798 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 51243 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Hajipour, Sepideh
  7. Abstract:
  8. The brain electrical signal (EEG) has been widely used in clinical and academic research, due to its ease of recording, non-invasiveness and precision. One of the applications can be emotion recognition from the brain's electrical signal. Generally, two types of parameters (Valence and Arousal) are used to determine the type of emotion, which, in turn, indicate "positive or negative" and "level of extroversion or excitement" for a specific emotion. The significance of emotion is determined by the effects of this phenomenon on daily tasks, especially in cases where the person is confronted with activities that require careful attention and concentration.In the emotion recognition problem, firstly, using proper emotion stimuli, different emotions are created for the subjects under study and the brain signals corresponding to each stimuli are recorded. The two main steps for solving of the emotion recognition problem are extracting suitable features and using appropriate classification or regression methods. In previous studies, different stimuli of vision and / or hearing have been used, and various linear and nonlinear features and classifiers have been investigated.In this thesis, the main goal was the improvement of linear regression algorithms to estimate the criteria for recognizing human emotions (including Valence, Arousal, etc.) in both matrix and tensor modes more efficient. For this purpose, we proposed three new matrix-based algorithms and two new tensor-based algorithms, which use useful information (including the sparseness of the mixing vector and the linear correlation of regression outputs) along with the linear regression cost function. The effectiveness of the proposed algorithms on simulated data has been investigated and their superiority to linear regression algorithms, PLS, HOPLS and LASSO was shown. Also, to apply the proposed algorithms on EEG data corresponding to emotions recognition DEAP dataset was used, and the AR coefficients were extracted from the EEG signals. The results obtained from the proposed algorithms in matrix and tensor modes were compared with those of the other linear regression algorithms, which in total showed the relative superiority of PLS-based methods compared to other linear regression methods
  9. Keywords:
  10. Emotion Recognition ; Linear Regression ; Electroencephalogram Signals Classification ; Tensor-based Algorithms ; Partial Least Squares (PLS)Regression ; Valence ; Arousal

 Digital Object List

 Bookmark

No TOC