Is Riemannian Geometry Better than Euclidean in Averaging Covariance Matrices for CSP-based Subject-Independent Classification of Motor Imagery?
Yassawe KainoldaBerdakh AbibullaevReza SameniAmin ZollanvariPublished in: EMBC (2021)
Keyphrases
- motor imagery
- covariance matrices
- lie group
- eeg signals
- riemannian manifolds
- euclidean space
- covariance matrix
- maximum likelihood
- riemannian metric
- pattern recognition
- log euclidean
- feature vectors
- machine learning
- distance measure
- support vector machine
- brain computer interface
- support vector machine svm
- gaussian mixture model
- classification accuracy
- geodesic distance
- eeg data
- vector space
- blind source separation
- shape analysis
- svm classifier
- model selection
- expectation maximization
- signal processing
- support vector